Printer Friendly

Text and time: a theory of testamentary obsolescence.

Events may occur after a will is executed that ordinarily give rise to changes of intent regarding the estate plan--yet the testator may take no action to revoke or amend the original will. Should such a will be given literal effect? When, if ever, should lawmakers intervene to update a will on the testator's behalf?

This is the problem of testamentary obsolescence. It reflects a fundamental, structural problem in law that can also crop up with regard to constitutions, statutes, and other performative texts, any one of which may become timeworn. This Article develops a theoretical framework for determining when lawmakers should--and should not--step in to revise wills that testators have left unaltered and endeavors to locate this framework in the context of other forms of textual obsolescence. The Article focuses on a variable denoted "friction"--that is, the extent of difficulty text makers face in revising texts on their own. Some testators become incapable of amending their wills, and some events display the dual property of altering testamentary intent while simultaneously disabling the testator from executing a new estate plan. In such instances, legal intervention to effectuate intent is warranted. Where testators remain at liberty to amend their wills following a change of circumstance, however, the case for legal intervention becomes uneasy. Nevertheless, lesser forms of friction may continue to operate, affording testators less practical opportunity to redo their wills, and hence again giving cause for lawmakers to interpret wills dynamically.

When they do act to reinterpret a will in light of changed circumstances, lawmakers should ordinarily follow the course that a majority of testators would choose, as default rule theory dictates. Yet, legal intervention to effectuate probable intent could implicate error costs, if testators believe a different rule is in effect. By matching default rules of will interpretation with common assumptions about what rules apply, lawmakers minimize error costs. As this Article demonstrates in the Appendix, under certain conditions an error-minimizing default is more efficient than a majoritarian default, a contribution to default rule theory.
TABLE OF CONTENTS

INTRODUCTION
I. INTERVENTION IN THE PRESENCE OF FRICTION
      A. Theoretical Prologue
      B. Loss of Capacity
      C. Loss of Opportunity
         1. Common Calamities
         2. Slayings
         3. The Relevance of Imputed Intent
II. INTERVENTION IN THE ABSENCE OF FRICTION
      A. Impossibility
      B. Other Changed Circumstances
         1. Error Costs
         2. Friction Redux
            a. Ripeness
            b. Lag
         3. Extrinsic Evidence
         4. The Code
CONCLUSION
APPENDIX


INTRODUCTION

Prior to the nineteenth century, Americans and Britons typically put off executing their wills until death was near. (1) The resulting estate plans were timely but not always tidy, for testators often conceived them in haste. One of the early arguments against freedom of testation in Great Britain was that testators "visited with sickness, in their extreme agonies and pains," might dispose of their estates "indiscreetly and unadvisedly." (2) Since the twentieth century, deathbed wills have grown comparatively rare, and as a consequence the risk of testamentary indiscretion has receded. (3) But every silver lining has its cloud. Wills drafted in the prime of life implicate a different peril--the risk of being overtaken by events. If a hiatus separates the time when a will is executed from the time when it matures, intervening occurrences--changes in the testator's life--may render it less well adapted to his or her subsequent circumstances.

This is the problem of testamentary obsolescence or, to borrow a scholar's turn of phrase, the "stale will." (4) Viewed structurally, it reflects a fundamental dilemma that recurs in our law. Whenever a court is called upon to apply the performative words of others, it must decide whether to read those words statically or dynamically, in spite of or in light of evolving facts. Time does its work, and contracts, statutes, and constitutions, inter alia, (5) along with wills, are all subject to the march of anno domini. Ultimately, lawmakers confront the core problem of textual obsolescence and ought to examine that problem in all fields of law from a common perspective, even if it yields different outcomes within individual fields. (6)

The replication of strategies to avoid textual obsolescence bears witness to the unity of the problem. Text makers themselves can update their words, of course, and codicils to wills stand beside statutory and constitutional amendments. (7) Alternatively, text makers can take immediate precautions against subsequent obsolescence. One possible response, developed independently within various textual categories, is to attach fuses to texts so that they self-destruct after a period of time. The texts will then enjoy too short a life span to become obsolescent. Thus, contracts typically cover "spot" transactions; long-term "relational" contracts are rarer. The statutory analogue is a sunset clause. (8) If Thomas Jefferson had had his way, the federal Constitution would have expired every nineteen years. (9) And some estate plans incorporate the same feature--a "conditional will" is tied to a looming hazard and becomes void if the testator survives that hazard. (10)

To the extent they can anticipate fortuities that would render a text anachronistic, text makers can also build into it preservatives against staleness. Contingency clauses often decorate wills and contracts. (11) Within some statutes, fallback provisions (usually anticipating the possibility of unconstitutionality) and indexing provisions perform an analogous function. (12)

Alternatively, text makers may concede the futility of trying to anticipate every contingency and empower a delegate to revise their texts as circumstances evolve. In effect, that is what legislators do when they incorporate standards into statutes; a court can then reinterpret their application over time. (13) In inheritance law, a power of appointment or discretionary trust serves this end. The donee of the power or the trustee will make distributive decisions as dictated by unfolding events.

The problem remains that text makers may decline or neglect to take any of these steps--a distinct possibility among the makers of testamentary texts. One estate planner offers a bleak assessment: "If truth were known, I believe we would be aghast at the number of outstanding wills of living persons in this country which are obsolete, as far as reflecting the present wishes of the testator." (14) When, if ever, should courts step in to update a text on its maker's behalf? Specifically in the realm of wills, should courts ever infer textual revisions that testators themselves never formalized in an executed writing?

In the pages following, we shall sketch a theoretical blueprint for analyzing this problem. And because similar issues arise in other textual arenas, we shall also strive to examine the problem in relation to its counterparts. In Part I, we consider situations where testators lack the opportunity to revise a will on their own, making judicial intervention potentially attractive. In Part II, we proceed to situations where testators do have that opportunity, making the case for judicial intervention more complicated and problematic. In the course of the analysis, and as elaborated in the Appendix, we shall propose a new sort of default rule, termed an error-minimizing default, that is demonstrably efficient under certain conditions and could prove to have broader theoretical significance.

I. INTERVENTION IN THE PRESENCE OF FRICTION

A. Theoretical Prologue

In theory, courts are constrained to abide by the meaning of a legally performative text, as a reflection of its authors' intent. Fidelity to the intent of legislators or framers follows from our observance of democratic principles. With regard to testamentary texts, lawmakers likewise acknowledge "the bedrock principle of honoring the intent of the testator," (15) a value grounded in public policy, if not political theory. (16) The problem of textual obsolescence arises, conceptually, when a legally performative text no longer corresponds with the sentiments of its authors or their successors-in-interest.

In a perfect world, the authors themselves would attend to the task of updating their texts with precision and dispatch. In the real world, that may not happen--and in some situations, we can anticipate the occurrence of obsolescence to be systematic. A key argument leveled against originalism in the context of constitutional interpretation is that the mechanism for amending the text is so cumbersome that the changed intent of the electorate (as successors to the framers) cannot realistically assert itself over time. (17) A similar argument can be put forward, albeit less forcefully, in the context of statutory interpretation. Here, the impediments to change are quantitative rather than qualitative: once the legal landscape becomes crowded with statutes--a pattern that sesquipedalian scholars have dubbed "statutorification"--legislators fight a losing battle to keep all of their acts current. (18) Inevitably, some acts are subordinated (an observation that brings us back to inheritance law, a field crawling with old vines). (19) Originalists may still object to judicial intervention to update a text as undemocratic. (20) Yet, if a court updates text by scrupulous reference to the sentiments of its authors, while ignoring the court's own sympathies or inclinations, it can still claim adherence to democratic (and testamentary) values.

In connection with testamentary texts, however, the assertion that obstacles stand in the way of currency is less compelling. Ordinarily, amending a will by codicil is simple and inexpensive; and whereas modern estate plans may combine multiple texts when disposing of assets outside of probate, multiplying in turn the number of required amendments, textual proliferation presents no parallel to statutorification. (21) For this reason, originalism remains the dominant perspective on interpretation of wills. "A will may be so easily revoked by the testator in his lifetime," one judge opined, "that the courts have been slow in permitting changes in circumstances to do by implication what the testator may so readily do for himself." (22)

That said, we can nonetheless identify settings in which this assumption breaks down, settings in which testators pursuing a path to will revision are bound to encounter "friction" of various kinds and degrees. Therein lie the rudiments of a case for dynamic interpretation of testamentary text.

B. Loss of Capacity

One circumstance that could stymie self-revision of a will is incapacity. Lawmakers deny incapacitated testators the right to alter their wills, lest they cease to display the detailed thoughtfulness that comprises one of the social benefits of granting freedom of testation in the first place. (23) Still, an unalterable estate plan may grow thoughtless anyway, given its unresponsiveness to the drift of events. All else being equal, the will of an incapacitated testator is frozen into place even more firmly than a constitution or a statute. (24)

Should lawmakers intervene to allow a guardian to update the incapacitated testator's will on his or her behalf? Considered structurally, the problem of incapacity is very like--indeed, an anticipatory extension of--the problem of dead hand control: to the extent a testator's will creates trusts that continue after death, we again face the prospect of an increasingly archaic estate plan that the testator is unavailable to amend. It is metaphysical to speak of effectuating intent ex post in either context: a testator who lacks capacity or dies has lost the capability to "intend," whether rationally or ontologically. (25) Still, lawmakers can undertake to effectuate probable intent ex ante. The question then becomes whether the typical testator would prefer that someone else step into his or her shoes upon incapacity or death to keep the estate plan current. Whereas some might appreciate this service, others would decline it for fear of abuse or want of insight into the testator's donative perspective. No polling data on testators' preferences in this regard have yet come to light.

Historically, guardians' tutelary powers failed to include a right to revise the estate plan of a ward. (26) Courts and legislators have gradually widened those powers, (27) a trend the Commissioners promulgating Uniform Acts seek to encourage. As originally drafted, the Uniform Probate Code (the "Code") stopped short of allowing a guardian to execute or amend a ward's will, but the Code did grant guardians power, with the court's approval, to execute or amend a revocable inter vivos trust. (28) Because such a trust supersedes a will, guardians have effective control over the estate plan. (29) In 1998, the Commissioners dropped the facade, revising the Code to allow guardians with the court's approval to make or amend a ward's will directly, (30) although few states have adopted this rule. (31) In ratifying or rejecting the guardian's testamentary decisions, the court "shall consider primarily the decision that the protected person would have made, to the extent that the decision can be ascertained." (32) An analogous provision appears in the Uniform Trust Code of 2000, giving the court power to update, "in accordance with the settlor's probable intention," the substantive terms of a trust otherwise immobilized by the dead hand where "circumstances not anticipated by the settler" intervene. (33)

There remains a structural contradiction within these respective provisions. Both ostensibly function to carry out benefactors' substantive intent. Yet both also take the form of mandatory rules. (34) In neither instance can benefactors opt out of the process for textual revision by express statement in the will or trust, (35) as they could if presented with a default rule. (36) Were these provisions framed as default rules, it would initially be incumbent on the Commissioners to determine that most benefactors would authorize a surrogate not selected by themselves to second guess their intent, a matter that calls for empirical inquiry. (37) For mandatory rules, no such inquiry is necessary, but then the Commissioners must justify the novel restrictions on freedom of testation that these provisions impose. (38) The wisdom of such restrictions is hardly manifest. (39)

C. Loss of Opportunity

Once we acknowledge that inability to update a will could provide grounds for legal intervention, we have to consider the range of situations where such a handicap could arise. Executing a will requires more than just mental capacity. It also requires time--always a scarce resource and sometimes very scarce.

It may happen that an event has the interesting, dual property of altering the testator's intent while also depriving the testator of a realistic opportunity to revise his or her estate plan to reflect the change. In that case, assuming they can readily deduce the testator's preferred revision, lawmakers further the cause of intent effectuation by amending the will on the testator's behalf. Although structurally analogous, the events falling into this category are situationally diverse.

1. Common Calamities

Consider the case where a testator and a beneficiary named under the will suffer a common calamity--for instance, an automobile accident in which one is the driver and the other a passenger--that mortally wounds both of them. Even if the beneficiary survives the testator by a short while, we can predict that the testator would prefer to substitute a different taker. Now on death's door, the beneficiary will have no occasion to enjoy the bequest, and it will pass in short order to others selected to inherit under the beneficiary's, instead of the testator's, estate plan. Yet, the testator has no chance to revise the will to take account of this change of circumstance. His or her own wounds, coeval with the beneficiary's, preclude execution of a codicil.

The first scholar to light upon this problem was an improbable one: John Wigmore, who made his reputation in the field of evidence, not wills. (40) Addressing the problem of simultaneous death, where there is no way of knowing who survived whom, and hence who is a beneficiary of whom, Wigmore observed that the question was properly one of intent, not of evidence. (41) This insight led him to endorse a rule of disinheritance for simultaneously dying beneficiaries, but also to propose--albeit as "a radical novelty" (42)--that the same rule extend to beneficiaries who survive the testator briefly. This proposal, Wigmore continued, was
   based on the idea that to change totally the devolution of an
   estate because of its vesting for a brief moment ... is contrary to
   the testator's intention, and therefore unfair.... [H]e certainly
   did not intend that it should go to ... a semi-corpse that retained
   only the semblance of physical vitality for a few moments. (43)


The drafters of the Uniform Simultaneous Death Act incorporated Wigmore's proposal into their provisional draft, (44) but they struck it from the final Act of 1940. (45) It was left to the drafters of the Uniform Probate Code of 1969 to rediscover the idea. Under the Code, all beneficiaries must survive the testator by at least five days in order to take their bequests, unless the will provides otherwise. (46)

Whether this approach is optimal remains a question. The difficulty with any bright-line postmortem survival (or "overlive") requirement is that it is arbitrary. (47) On the other hand, a variable standard (for instance, a rule disinheriting any beneficiary who dies as a result of a common accident with the testator) is flexible but would breed litigation. Estate planners routinely include overlive requirements in the wills they prepare, and these give some indication of the typical testator's intent. Estate planners avoid framing these requirements as variable standards, (48) but they usually run longer than the five days provided for under the Code--thirty or sixty days are now commonplace. (49) Medical technology has progressed since the Code was promulgated, and "semi-corpses" can linger longer than in Wigmore's day. In its treatment of this segment of the problem of obsolescent wills, the Code may itself have grown to become obsolescent. (50)

2. Slayings

Another case presenting the same structural circumstance differs in every other respect. Consider a beneficiary who slays the testator. Pointing a sharp metal object at the testator's throat and thrusting it forward is the sort of act likely to snap the sociological bonds that previously tied the testator to his or her assailant. But the testator lacks time to communicate the change of intent following from that act: in this instance, the sword is mightier than the pen. For lawmakers to intervene here (as with a common calamity (51)) is to impute a change of intent that did come about, by hypothesis, and which empirical data may confirm, (52) but that the testator had no opportunity to express.

However judiciously, lawmakers have responded to the problem. Under "slayer statutes," currently found in forty-four states and also featured in the Uniform Probate Code, slayer-beneficiaries forfeit their inheritances. (53) Historically, though, lawmakers have propounded a different rationale for this outcome: "the prevention of unjust enrichment, in accord with the maxim that a wrongdoer cannot profit by his or her wrong." (54) In other words, the rule has criminological overtones and shares the same inspiration as Son of Sam laws. (55) The rule's alternative justification as a means of effectuating testators' probable wishes has not gone unnoticed, (56) but it is plainly subordinate. Thus, slayer statutes set out a mandatory rule that supersedes even an express provision to the contrary in a will (as conceivably might appear in connection with an assisted suicide). (57)

This much is reasonable enough. Where the policies of criminal law and inheritance law collide, avoiding wrongdoing should take precedence. (58) The troubling structural aspect of slayer statutes is their singular fixation on wrongdoing, which narrows their scope. The policies of unjust enrichment-cum-deterrence on one hand, and intent effectuation on the other, cover intersecting sets of circumstances. The union of these two sets defines the proper scope of slayer statutes.

Consider the matter of criminal defenses. Slayer statutes only invalidate a bequest to one "who feloniously and intentionally kills the [testator]." (59) Hence, with rare exception, they do permit a slayer-beneficiary who successfully pleads insanity to take under the will. (60) Confining the operation of the statute in this way tracks criminal law policies. From the vantage of inheritance policy, however, such confinement is doubtful. A slayer-beneficiary's subjective inability to appreciate the wrongfulness of his or her conduct may not typically forestall the victim's last-minute change of intent. Likewise, and without exception, slayer statutes permit a slayer-beneficiary who successfully claims self-defense to take under a will. (61) Yet, in these circumstances, we can infer a change of intent with some confidence. If, after all, evidence shows that a slayer-beneficiary was defending against the testator's deadly force, as required for the defense to succeed, (62) then that same evidence indicates that the testator's relationship with the beneficiary had grown, to say the least, discordant. Ironically, what is exculpatory for one purpose is probative for the other. Slayer statutes should reflect probable changes of intent in both of these situations, (63) but only as default rules (since criminal-law policy is no longer implicated).

Unintentional killings, which come under the heading of involuntary manslaughter, in most jurisdictions fall outside the purview of slayer statutes. (64) As a criminological matter, broadening their reach would serve to deter the aggravated negligence required for a finding of involuntary manslaughter, (65) a policy lawmakers pursue elsewhere by assessing punitive damages, another "quasi-criminal" (66) remedy. Be this as it may, most testators might prefer to disinherit slayers in these cases-particularly if willful conduct to injure the testator short of death, but nevertheless resulting in death, was involved. (67)

Except in two jurisdictions, (68) beneficiaries who commit acts of violence against testators that do not cause their deaths remain entitled to inherit under their wills. Because they do not accelerate an estate plan, injurious assaults afford no wrongful gain to an assailant-beneficiary. Nor, from the standpoint of intent effectuation, do injurious assaults disable testators from retaliating by revising their wills.

We can hypothesize one special scenario, however, implicating just such a disability. Consider an assault that injures a testator seriously enough to rob him or her of testamentary capacity. Even though the

testator may live on for years, he or she now lacks the wherewithal to respond by amending the will. (69) If polling data confirms that most testators would disinherit their assailants under these conditions, then slayer statutes should again cover the case with a default rule.

3. The Relevance of Imputed Intent

Of course, the deprivation of opportunity to amend a will in connection with incapacity, common calamity, and slaying only arises at the time when the disabling event occurs. A testator could plan ahead for any such predicament by inserting contingency clauses into his or her will. Because this opportunity exists before the fact, we might argue, lawmakers have less reason to concern themselves with effectuating intent thereafter. Fidelity to intent follows from our embrace of freedom of testation, yet its value to society derives from the scope this freedom creates for active estate planning, not from reconstructing presumptive shifts of intent. On this basis, we might temper our concern for intent effectuation, placing the onus on testators to expound their wishes--whether present or anticipated--within an executed writing. (70)

Yet even if effectuating imputed intent under conditions of friction goes beyond the pale (and policies) of freedom of testation, we have other reasons for doing so. This move promotes both efficiency and equity. It is expensive for testators to plan for every contingency, however remote. Whereas in most instances testators can wait to see whether unlikely contingencies come to pass, and then react with codicils, they do not enjoy that luxury when friction would intervene. Intent-effectuating default rules that cover disabling events spare testators transaction costs. (71) What is more, human foresight is imperfect. The cognitive costs of simulating future states of the world (known in the psychological literature as prospection) produce simplified visions that overlook possibilities and thereby condemn testators to some degree of textual incompleteness. (72) Persons also tend selectively to put out of their minds (via denial) prospects that are unpleasant to contemplate. (73) Estate planners have encountered all of this in practice, observing in clients an inclination to view their world as static and to resist pondering painful contingencies. (74)

Professional will drafters better understand what the future may bring, and their work products could--if need be--take into account contingencies that are coupled with disabling events. Estate planners routinely provide for the contingency of a common calamity, (75) and some also attend to the possibility of incapacity. (76) But a significant fraction of wills are composed by testators acting without benefit of counsel, (77) and these lay drafters are poorly equipped for the task at hand. By establishing a panoply of intent-effectuating default rules to cover the scenarios addressed in the foregoing pages, lawmakers can provide testators with a substitute for counsel they may lack the means to afford. In so doing, lawmakers help to destratify the process of will-making. (78)

Ultimately, then, intent effectuation holds value in itself, not merely as a concomitant to freedom of testation.

II. INTERVENTION IN THE ABSENCE OF FRICTION

A. Impossibility

When we turn to changes of circumstance that testators remain at liberty to answer by revising their wills, the case for legal activism to update text becomes uneasy. Yet, where the change is of a nature as to make the original estate plan impossible to implement, some sort of intervention has to occur. Courts can do many things, but they cannot do the impossible. Here they have no choice but to deviate from the strict letter of a document's text.

In the context of legislation, the traditional canon of construction holds that when statutes are or become impossible to carry out (constitutionality being the usual stumbling block), they are invalid. (79) A court will not, in other words, conjecture how legislators would rather proceed within the realm of the possible, presumably because they can make that decision for themselves once the status quo has been restored. Testators can also revise an impossible estate plan; but when a court encounters one that a decedent failed to revise, it is too late to restore the status quo. At this point, the court can only substitute a new plan for the one stymied by events. (80)

Impossibility in the context of wills arises chiefly in two situations: (1) where property testators bequeathed no longer remains within their inventory of possessions and hence is no longer theirs to give away; and (2) where named beneficiaries (by analogy) are no longer alive and hence are unavailable to accept bequests. By striving to effectuate probable intent in these situations, lawmakers can again destratify the inheritance process and reduce transaction costs. (81)

Determining how best to accomplish this result requires analysis at two levels. First, we must ask how testators in general would prefer to revise an impossible estate plan. Would the typical testator wish to substitute for a bequest of specific property no longer owned a bequest of other property equal in value? Would the typical testator wish to substitute for a bequest to a predeceasing beneficiary a bequest to that beneficiary's children, or perhaps spouse, as opposed to the testator's residuary legatee (or heirs)? These questions demand empirical inquiry.

Second, we must address whether evidence of a testator's intent in the particular case should override assumptions of intent drawn in the general case. Surely, courts should bar evidence of the testator's declarations following, and in response to, the occurrence of impossibility. No friction appears in these scenarios, so testators remain free to respond by formalizing any new intent that impossibility forces them to contemplate. Lawmakers should require formalization of changed intent, at least in the absence of friction, for the same reasons they require formal execution of the original will: it certifies finality of intent and offers a measure of protection against fraud. (82) At the same time, one could make a case for allowing courts to consider other features of the estate plan, together with extrinsic evidence of a testator's underlying motives for the impossible bequest, and of the precise circumstances causing impossibility, in order to glean how that particular testator would have provided for the contingency ex ante. Case-by-case analysis can effectuate intent (which may vary a great deal) more accurately, and reading between the lines of the original will does not strictly violate the execution requirement. The other side of the coin is that case-by-case analysis implicates additional litigation, and thus administrative costs. (83) Furthermore, the analysis involves inference rather than construction, since the testator never expressed an executed intent concerning unanticipated contingencies. Hence, in seeking to divine how a testator would have answered a question that was never formally posed, the court could still fall victim to error, if not fraud. (84)

Existing rules covering these matters leave much to be desired. Consider first bequests to predeceasing beneficiaries. These bequests "lapse" and flow into the testator's residuary estate under common law, unless an "antilapse" statute diverts them to the predeceasing beneficiary's descendants. In a plurality of states, antilapse statutes supersede the common law of lapse only as concerns bequests to close blood relatives. (85) The underlying assumption is that testators' relationships with their kinfolk extend to their descendants. If blood relatives predecease, the typical testator would likely wish to provide for their progeny instead. On the other hand, testators might have weaker ties to the families of beneficiaries who are not blood relatives (employees, for example), and so we can have less confidence that the typical testator would wish to treat their progeny as testamentary surrogates. (86)

Perhaps, perhaps. But all of this needs empirical testing. It might turn out, for instance, that with regard to some categories of beneficiary the typical testator would prefer to name as a substitute taker the beneficiary's surviving spouse--some extant data raises this possibility. (87) And we also need to discover where to draw the line (assuming one should be drawn at all (88)) between those bequests that lapse and those that do not. All told, depending on the state, these lines currently fall all over the substantive map. (89) The rule found in the Uniform Probate Code has its own unique features, but if the Commissioners wish to separate their line from this confused jumble, they must support it with something more concrete than inference. (90)

When we turn to the structurally analogous problem of specifically bequeathed property that disappears before the testator's death, the legal landscape grows bleaker still. These bequests automatically "adeem by extinction"--that is, they become void, and the named beneficiary receives nothing. (91) This rule is not based on empirical evidence or even inference of intent but follows rather from the formalistic finding of impossibility--the "fact that the thing bequeathed does not exist." (92)

Lawmakers nevertheless could replace the bequeathed property with similar property or funds of equivalent value. Doing so might or might not correspond with the typical testator's intent. Lawmakers inclined to explore the question could study the empirical significance of any number of variables: the kind of property involved, (93) its relative value, the relationship between the testator and the beneficiary, (94) the circumstances that caused the property to disappear, (95) or what followed. (96) The Uniform Probate Code takes a stab at the problem, carving out a number of exceptions from the traditional doctrine of ademption by extinction, but again without the guidance of empirical data. (97)

In most jurisdictions, the rules of lapse and ademption take effect unless overridden by contingency clauses in the will itself. Other evidence of probable intent is inadmissible. (98) Yet lawmakers have not followed this course with spotless consistency: in one other (relatively rare) instance of impossibility, they do admit other evidence. If a testator makes a conditional bequest, and the condition proves impossible to carry out, "impossibility ... excuses lack of performance if, and only if, this result is the appropriately ascertained intent" of the testator. (99) In this situation, the court is free to "ascertain" intent by examining extrinsic and other intrinsic evidence. (100)

But for its rarity, the problem of impossible conditions is structurally analogous to lapse and ademption. Here again, "transferors do not have in mind the contingency that impossibility may prevent ... performance," leading us to inquire hypothetically "what [their] intent would have been had [they] ... anticipated the impossibility." (101) Whether the greater likelihood (but not assurance) of intent effectuation via case-by-case assessment of evidence merits the administrative cost is hard to say. The easier point to make is that lawmakers ought to treat all of these problems symmetrically.

B. Other Changed Circumstances

We are left, finally, with changed circumstances that neither disable the testator from revising the will nor render the estate plan impossible by its terms to carry out. Whether lawmakers should update an estate plan by implication of such circumstances when the testator has not lifted a finger to amend the will is ultimately the most interesting--and most neglected--question we have before us. And, theoretically, we can imagine a range of responses.

The formalistic move is to give effect to the literal terms of the will. Circumstances change every day of our lives. Lawmakers may reasonably require testators to execute a codicil to reflect a change of mind, at least when nothing stands in the way of their doing so. Insistence on formal execution of revisions eases the task of verifying the substance and finality of testamentary intent. At any rate, if testators have gone to the trouble of executing an initial will, lawmakers have some cause to assume that testators will update the will if and when it suits them to do so. (102)

The alternative antiformalistic move would allow courts to weigh the effects that changed circumstances have on a testator's intent and to implement that reading even in want of a codicil. The virtue of this approach is its flexibility "amidst the endless variety of human affairs," (103) recognizing that, for whatever reason, testators may lose sight of their estate plans. Its vice, though, is the danger that courts will fail accurately to reconstruct intent due to error or fraud, coupled with the potential administrative cost when every will is subject to such reconstruction.

As a matter of law, nearly all jurisdictions reject the antiformalistic move. (104) "[A] court may [not] wander from the actual words of a will into the region of conjecture as to what it is reasonable to suppose the testatrix would have done had she contemplated a certain event happening," (105) one state supreme court opined, emphasizing the uncertainty of the enterprise: "A court is not free to roam such unfenced fields of speculation." (106) Other courts have deplored the idea. (107)

Still and all, no jurisdiction follows the formalistic move scrupulously either. At least after erecting fences, lawmakers have ventured out into the field of speculation, amending an estate plan in discrete situations that vary from state to state. Three triggering events predominate: where the execution of a will is followed by divorce, by marriage, or by childbirth. (108) In most jurisdictions, if a testator makes a will and subsequently marries, the new spouse receives an intestate share under the premarital will; likewise, if a testator makes a will and subsequently gets divorced, the former spouse loses his or her bequest under the predissolution will; finally, if a testator makes a will and subsequently becomes a parent, the new child receives an intestate share under the prenatal will. (109)

These rules effect what is known, rather deceptively, as implied revocation (or alternatively, revocation by operation of law). (110) In truth, the testator performs no action that is intended, implicitly or explicitly, to revoke the estate plan in these situations; lawmakers simply reckon that these dramatic changes of circumstance will likely precipitate a shift of testamentary intent. Nearly two centuries ago, Chancellor Kent submitted that implied revocations "are founded upon the reasonable presumption of an alteration of the testator's mind, arising from circumstances since the making of the will, producing a change in his previous obligations and duties." (111) Now, as then, courts defend the doctrine of implied revocation as "anticipat[ing] that, upon undergoing a fundamental change in family composition ... [testators] would most likely intend to provide for their new family members, and/or revoke prior provisions made for their exspouses." (112)

Implied revocation nevertheless has an underside. Even if we can establish the empirical likelihood that testators would respond to a particular change of circumstance with a particular revision of the estate plan, we cannot surmise that amending a will on its author's behalf effectuates intent with the same likelihood. The issue is not whether someone would probably want to revise a will following a consequential event, but whether someone who has not done so would probably want to do so. And that, on reflection, is a separate question. (113)

1. Error Costs

Consider the case of a testator who executes a will in favor of his or her spouse and the two subsequently divorce. If we decide whether to leave standing or revise the will on the basis of the likelihood that the testator's intent changed, we minimize transaction costs. But there exists a second consideration, which has in fact loomed larger in judicial discussions. If lawmakers leave standing a will that most testators intend to revise, some will still neglect to act, out of inadvertence or procrastination. (114) These testators bear not a transaction cost, but an error cost. On the other hand, if lawmakers invoke the doctrine of implied revocation to amend the will, other testators may leave it in place not because they are happy with the doctrine as applied, but because they are ignorant of what legal effect divorce had on the estate plan, mistakenly assuming it remains in effect as before, and intending that result. These testators also bear an error cost. If legal ignorance of this sort is more common than procrastination, we minimize error costs by imposing the default rule not that most testators prefer, but that most presuppose. This we may dub an error-minimizing default.

Orthodox default rule theory posits that majoritarian default rules are efficient; but orthodox theory, which developed in connection with commercial transactions, assumes that contracting parties are repeat players who are informed (or can efficiently become informed) about the applicable default rule. (115) Information costs pose a more serious obstacle for testators, for whom estate planning is a one-time (or at best infrequent) activity. (116) What is more, in situations where testators already have an estate plan in place, and where the consequential event does not prompt consultation with counsel, they may not even be aware of the need to become informed--they may simply assume that the estate plan they formally executed remains in effect until they see fit to change it, thereby magnifying information costs. (117) In the Appendix, below, we develop an economic model to prove the potential efficiency of error-minimizing defaults, and to ascertain the conditions that must exist for them to achieve greater efficiency than majoritarian defaults--a possibility not previously recognized or explored in the default rule literature. (118)

And so we face a quandary. Whether lawmakers take it upon themselves following a consequential event to revise a testator's will or leave it be, the intent of some testators will be frustrated--and without data, we cannot predict which option is the lesser evil. Nor is this an easy matter to investigate empirically. All of the extant studies of inheritance patterns and preferences rely either on random surveys (asking either factual or hypothetical questions) or on probate records. These can serve to illuminate probable intent, but they are uninformative on the distribution of error costs. The only way to investigate those costs is to develop a data set of actual testators who retain their original wills following consequential events of a particular sort, and then to poll testators concerning why they have left their wills unchanged. In point of fact, no such study has ever been conducted. The doctrine of implied revocation rests upon a flimsy foundation of evidence and analysis.

2. Friction Redux

There exists, however, an alternative approach to the problem at hand. That is to acknowledge a doctrine of implied revocation but confine it to cases where we have less room for doubt that a testator's failure to amend his or her estate plan following a consequential event was unintentional.

Either of two factors, structurally speaking, can retard the process of testation and thereby reintroduce friction into the system--not disabling friction, to be sure, but enough to make responsiveness to changed conditions unlikely. These factors we may label, respectively, ripeness and lag.

a. Ripeness

Testators may intend to amend an estate plan but nevertheless fail to do so, because the time is not yet ripe. We can identify several situations where estate planning might appear premature even though testamentary intent has already changed. The clearest case is conception. A parent-to-be may wish to bequeath to an unborn child. Still, the parent might postpone revision of the estate plan until the child is born and named. Were the parent to die unexpectedly prior to the child's birth, or during childbirth, the parent's failure to amend the will could trace to this variant of friction. Of course, a testator can anticipate childbirth by creating a class gift, or by leaving a bequest to an embryonic child. (119) But a testator's failure to consider the risk of premature death is entirely plausible, psychologically. (120) In this circumstance, a stronger case appears for an implied revision, assuming it conforms to the typical testator's intent. (121)

Another occurrence potentially causing testamentary intent to change, yet where estate planning might seem premature, is marital engagement. Persons whose relationships are sufficiently close that they plan marriage may already intend to bequeath something to each other. (122) Whereas nothing prevents fiances from amending their wills immediately, this transitional phase of the relationship might seem too fleeting to merit planning. The couple might understandably wait until the main event, disregarding the possibility of a premarital death. (123) Here again, we have cause to apply a doctrine of implied revision in accord with typical intent (once established), although no jurisdiction has yet pioneered such a rule.

A third scenario raising the problem of ripeness is more complex and requires elaboration. Consider the case of a testator who makes an inter vivos gift to a beneficiary who is also designated to receive a bequest under the testator's will. If the bequest comprises a specific item which the testator gifts to the beneficiary inter vivos, then it is adeemed by extinction. (124) But suppose the bequest comprises a general amount of cash. Is the gift given in lieu of the bequest and hence subtracted from the amount left to the beneficiary at death (thereby accelerating distribution)? (125) Or is the gift independent of the bequest and hence added to the amount the beneficiary receives at death?

At common law, the applicable doctrine is known as ademption by satisfaction. It provides that whether a gift is held to satisfy or augment a bequest depends on the testator's intent, as gleaned from declarations contemporaneous with the gift. (126) Whereas a rebuttable presumption of intent to satisfy bequests applies to children, a rebuttable presumption of intent to augment bequests covers all other beneficiaries. (127) Under the Uniform Probate Code, however, a presumption of intent to augment applies to all beneficiaries and is rebuttable only by a contemporaneous writing by the benefactor or a written acknowledgment by the beneficiary. (128)

On first sight, either approach appears anomalous. If we conceptualize a gift as just another event occurring during the hiatus between the execution of an estate plan and death, then declarations of intent not expressed in the will should be inadmissible, just as in connection with the other intervening events already discussed. (129) What makes the case different, however, is that a gift is not a mere change of circumstance, but also a gratuitous transfer--that is, a legally performative act. Whereas lawmakers are ordinarily chary of unformalized declarations of intent, given the uncertainty of their purposefulness and finality, when they accompany a legally performative act--here, formal delivery of a gift-declarations become more credible. Hence, lawmakers allow donors orally to declare the terms of a gift (viz. whether it is temporary or permanent, conditional or unconditional) when donors deliver them. (130) Ademption by satisfaction represents an extension of this policy.

What presumptions should apply when benefactors make gifts to will beneficiaries is a question as yet unexplored in any empirical study. The Commissioners claim that "[m]ost inter-vivos transfers today are intended to be absolute gifts or are carefully integrated into a total estate plan." (131) Is that a fact? No references accompany the assertion, so it would appear to be mere conjecture--one, incidentally, that commentators dispute. (132)

The Commissioners add that "[i]f the donor intends that any transfer during the donor's lifetime be deducted from the donee's share of his estate, the donor may ... execute a will so providing." (133) Yet there remains one situation where such action is subject to friction. Consider gifts given upon milestones. A parent may make a practice of giving substantial gifts to his or her children upon their attaining a certain age, or upon marriage, or upon the launching of their careers. If all goes to plan, they will receive equal treatment once the last of the children reaches the milestone--before that time, the gratuitous transfer to the last child is premature. But what if the parent dies ahead of the milestone? Of course, the parent could provide for that possibility under a will, but he or she may again fail to anticipate an untimely death.

The Commissioners appear to have given no thought to this scenario. It did not escape another body of lawmakers who set their minds to the problem a bit earlier in time. Under the Code of Hammurabi--the Babylonian lex scripta dating to the third millennium BC--we find that
   [i]f a man has made a gift ... to his son ... then, after the
   father has gone to his fate, when the brothers divide he shall
   retain the father's present which he has given him over and above
   the equal part that he shares in the possessions of his paternal
   house. (134)


In other words, the Babylonians--like the Commissioners--declined to apply a doctrine of ademption by satisfaction to children. But that was not the end of the matter:
   If a man has married wives to the children he has had, but has not
   married a wife to an infant son; then, after the father has gone to
   his fate, when the brethren share the possessions of the paternal
   house, they shall give silver for a bride-price to their infant
   brother who has not married a wife, besides his share; and he shall
   be married to a wife. (135)


And so, the ancient Babylonians offered up a refinement of the doctrine of ademption by satisfaction that we should hearken to today--along with a delightful reminder that the problems of inheritance law are well and truly eternal ones.

b. Lag

A more general problem arises out of friction that operates to slow, rather than to halt, the estate planning process. Testators do not revise their wills overnight. They may react sluggishly to changed circumstances. If they die nearly simultaneously with consequential events, as we have seen, friction becomes overwhelming. But even if they die shortly thereafter, their opportunity to respond to changed circumstances is limited, and the probability that their unresponsiveness was inadvertent increases. With each passing day, however, this risk recedes, while the likelihood that they left their wills unchanged on purpose creeps up commensurately.

This observation suggests a variation on the theme of implied revocation. Lawmakers could give effect to the doctrine but confine it to an interval of time following the consequential event--hence creating only a temporary presumption. Temporary presumptions comprise an unusual structural form in our law, although hardly an unprecedented one. Similar in spirit is a rule found in the Uniform Probate Code rendering ademption by extinction inapplicable to bequests of property "sold ... by a conservator ... for an incapacitated principal ... [unless] it was adjudicated that the testator's incapacity ceased and the testator survived the adjudication by one year." (136) The longer the testator who has recovered capacity fails to respond to the conservator's action by amending the estate plan, the stronger the inference that the testator approves of the consequences. (137) Determining the proper duration for an implied revocation depends on a finding of the typical lag in estate planning following consequential events, a space of time that could vary with the state of health of the testator, (138) or with the sort of event at issue, depending on how distracting or debilitating it tends typically to become, (139) and on how distantly it can be anticipated. (140)

The idea of temporally defining the doctrine of implied revocation has surfaced now and again in the past. In his sixteenth-century treatise on wills, Henry Swinburne hinted at a caveat to what was then the general rule that wills executed prior to childbirth remained effective as written:
   And albeit the testator after the making of the testament have a
   childe borne unto him, I suppose that the testament is not presumed
   thereby to be revoked, especially if the testator did live a long
   time after the birth of the childe, and might have revoked the
   testament, and did not. (141)


The idea has also flickered in the American case law, appearing primarily in dicta. In a case of first impression in Minnesota, the high court of the state had to determine the effect of divorce on a testator's predissolution will. In this instance, the testator had died just thirty days after the divorce was granted. In holding the bequest to the former spouse void, the court cabined its ruling:
   Whether the fact that the testator in such a case permits his will
   to remain unchanged ... for a number of years after the divorce and
   settlement would militate against the conclusion of implied
   revocation we need not determine. Testator in this case died within
   30 days after the settlement, and his failure expressly to [re]voke
   within that time certainly creates no inference that he intended
   his will to continue in force. (142)


Around the same time, the high court of Pennsylvania faced the issue raised hypothetically in the Minnesota case and concluded along the same lines:
   The divorce was granted eight years before [the testator's] death
   .... He had ample time and opportunity to consider ... [changing]
   the [life insurance] policy, but death came to him without his
   having made a change in the beneficiary ... . Under the
   circumstances it is evident that he never intended or desired to
   exercise his power. (143)


Other American courts have adverted to this distinction, (144) but it never crystallized into a rule in any state until 1984. In that year, the high court of Washington held that it would presume an intent to revoke a predissolution life insurance policy naming the divorced spouse as beneficiary if the dissolution decree indicated an intent to divest the former spouse of his or her rights as beneficiary, and if that intention was
   acted upon within a reasonable time after dissolution by formal
   execution of the change of beneficiary. If the intention is not
   acted upon within a reasonable time, the owner should be deemed to
   have decided to retain the named beneficiary as the one entitled to
   the proceeds. (145)


Reasonability of lag time under this ruling could vary with the circumstances, but in no event could it exceed one year. (146)

This rule died on the vine. The opinion was superseded in a matter of years by a statutory enactment of the more traditional, time-independent sort. (147) Nevertheless, the rule, or some variation of it, merits serious contemplation and it could serve as a model in other situations as well. Newlywed testators, by extension, cannot be expected to revise their estate plans immediately--but once the sunsets of the honeymoon are over, the assumption of changed intent could also sunset. So long as lawmakers confine implied revocation to a short interval of time around the consequential event, they need not concern themselves with the empirical complexities associated with error costs; (148) they need simply to identify those situations where an event is statistically likely to transform intent.

Indeed, empirical inquiry might well reveal a host of triggering events beyond the ones currently covered in most states by rules of implied revocation. (149)

Alternatively, one could conceive a limited doctrine of implied revocation framed as a standard, granting courts power to reform any will in accord with the testator's probable intent where a testator lacked a reasonable opportunity, in light of the facts, to update the estate plan in response to a significant change of circumstances. (150) Such a power would solve the problem at hand in the most flexible way, allowing courts to consider every possible change of circumstance, which lawmakers cannot anticipate comprehensively, along with the varying frictions impeding vel non individual testators. So long as the doctrine only operates temporarily, it should produce less litigation and narrow the scope for judicial intervention. At the same time, such a standard would free courts from the disciplines of text and probability, raising the risk of judicial error. If most testators would prefer to avoid that risk, then the principal effect of such a rule would be to cause testators to override it by so providing in their wills, pointlessly increasing transaction costs. (151)

3. Extrinsic Evidence

Another issue raised in connection with presumptions of implied revocation is whether extrinsic evidence should be admissible to rebut them. Jurisdictions have long differed on this question.

The policies at stake here are indistinguishable from those concerning instances of impossibility. To honor a testator's declarations of intent under these conditions would undermine the requirements of will formality. (152) On the other hand, extrinsic evidence can cast light on a testator's preferences, even if it cannot illuminate them with crystal clarity. None of the events presently triggering implied revocation has so unequivocal an impact on intent that extrinsic evidence is superfluous. (153) If lawmakers wish to minimize the litigation costs associated with admitting extrinsic evidence, as some opinions suggest, (154) then the applicable statutes could identify specific badges of probable intent sufficient to overcome an inference of revocation--just as the statutes fix upon specific, extrinsic facts to create inferences of revocation in the first place--although these again would require empirical verification. (155)

From the existing cases we can distill a number of recurring themes that hold promise as badges of probable intent. To wit: Was the will the testator executed prior to a consequential event made long before it occurred or near enough to it to allow the testator to anticipate the event when he or she formulated the estate plan? (156) Following the consequential event, how did the testator treat the will? Was it discovered among the testator's important papers, or among worthless papers? (157) Specifically in the context of divorce, does the former spouse have custody over minor children of the testator? (158) And also in that context, what was the de facto relationship of the testator and the former spouse following divorce? In a surprising number of published cases, the two have continued to live together as a couple. (159)

4. The Code

We turn finally to assay the Uniform Probate Code provisions on implied revocation. These cover the conventional trio of childbirth, marriage, and divorce as triggering events and operate (again according to tradition) irrespective of the length of time a testator leaves the will unchanged. In other respects, though, the Code features eye-catching, if not always clear-sighted, innovations.

To their credit, the Commissioners have led the way in applying empirical evidence to implied revocation in one instance. The Code pares back the doctrine in connection with wills executed before childbirth in deference to studies showing that modern wills typically place the surviving spouse ahead of children in the scheme of inheritance. (160) Other doctrinal novelties, however, want such a statistical foundation; the Commissioners have not shied away from translating into action their own suppositions about how the average testator responds to events.

Consider the Code's provision on wills executed before marriage. The Code impliedly revokes the premarital will to the extent necessary to furnish the new spouse an intestate share but excepts from revocation any bequests to children or their descendants under the premarital will. (161) Although the comment is silent, the Reporter elsewhere defends this refinement as directed to "late-in-life marriage in which [the testator's] instincts would likely be to want to continue to provide for her children from her first marriage." (162) Apparently, the Reporter relied on his own judgment to establish those instincts--no citation to evidence follows the pronouncement. (163)

One wonders whether the Reporter's assessment would change were he to consider that this exception to the doctrine of implied revocation under the Code covers a broader assortment of cases than just late-in-life remarriage. The second marriage could come early in life, following rapid divorce (a far more common phenomenon, statistically (164)) or the premature death of the first spouse. The child or children benefitting under the premarital will could still be minors, and the testator (with a new spouse in the household) might have custody over them. And the testator could have additional children with the new spouse. (165) At any rate, some data on testamentary preferences following remarriage are extant. All studies have found a more prevalent intent to provide for the new spouse than the Reporter infers. (166)

In this instance, the Commissioners' intuitions led them to a more parsimonious application of the doctrine of implied revocation. In another instance, however, they have chosen to expand it. The Code's provision revoking bequests under a will executed prior to a divorce covers not only bequests to the former spouse, but also bequests to any relative of the former spouse who is not independently related to the testator. (167) The Commissioners justify this rule in the accompanying comment:
   Given that, during the divorce process or in the aftermath of the
   divorce, the former spouse's relatives are likely to side with the
   former spouse, breaking down or weakening any former ties that may
   previously have developed between the [testator] and the former
   spouse's relatives, seldom would the [testator] have favored [their
   inheritance]. (168)


Accepting the Commissioners' reasoning for the sake of argument, we may observe that the rule they crafted in response fails to operate reciprocally. Whereas bequests from a testator to a former spouse's relatives are revoked by implication upon divorce, bequests from those relatives to their former son- or daughter-in-law (for example) remain effective under the Code. (169) Yet, if they "side[d]"170 with their consanguine in the divorce, they should be equally intent on disinheriting their former affine. This doctrinal asymmetry represents a logical flaw in the Code.

As for the reasoning itself, it appears in the comment as an ipse dixit, unsupported by a single scholarly reference. Instead, the Reporter elsewhere defends the rule with a hypothetical:
   Ben and Elaine had a happy ... marriage. [Ben's] will ...
   designated Elaine as the primary beneficiary at Ben's death. Ben,
   an only child whose parents had died, attempted to foster good
   relations with Elaine's parents by naming them as the alternative
   beneficiaries.... Eventually Ben and Elaine[] ... divorce[d].
   Shortly after the divorce, Elaine married Carl. And shortly
   [there]after ... Ben died.... If Elaine does not benefit under
   Ben's [predissolution will], who does? Conventional statutes
   resolve that question by invoking the fiction that Elaine
   predeceased Ben. In consequence, Ben's property would apparently
   pass to Elaine's parents as alternative beneficiaries. Because such
   an outcome appears inconsistent with Ben's likely intent, the 1990
   UPC also revokes any disposition to the former spouse's relatives.
   (171)


Of course, one can always construct a scenario for which a rule yields an appropriate outcome. That is the forensic equivalent of drawing a bull's-eye after one fires the rifle shot. (172) The real question is one of statistical probability within the ambit of possible scenarios. The bequest at issue may designate an alternative beneficiary to the former spouse. Contrarily, the bequest may be in addition to the one for the former spouse. (173) The beneficiary named may be an ascendant or collateral affine. Or--probably in a large majority of instances where this rule comes into play--the beneficiary may be a stepchild of the testator. (174)

On the specific matter of inheritance patterns following divorce, we lack hard data. Nonetheless, the sociological literature offers an abundance of material on divorce, stepfamilies, and (less so) redivorce that is germane to the issue at hand, including hundreds of published empirical studies. (175) The Commissioners failed to mine this body of research--they did not do the leg work. Had they delved into the literature, it might have given them pause.

The question is whether a spouse's relationship with stepchildren and other quasi-kin (in the sociologists' jargon) depends on the continuity of the marriage or instead can "take on a life of its own." (176) Empirical research on supportive relationships between a child-in-law and parent-in-law finds that these can go on after divorce, although they do so relatively infrequently. (177) The two empirical studies of testation by former in-laws published to date explore attitudes rather than actual behavior and focus on bequests to members of a divorcing couple. The studies found only a minority of subjects to consider bequests to a former son- or daughter-in-law appropriate. (178) Still, we should interpret these data cautiously since the studies found greater approval of bequests to former children-in-laws if the testator's social relationship with them survives a divorce, (179) and the likelihood of that occurring is tied to their closeness during the marriage. (180) When wills include bequests to in-laws, that fact in itself could signify that the parties were close to begin with, although we have no assurance of such a correlation.

Turning to stepchildren, we again face a shortage of hard data but a surfeit of related data. Hardly any empirical studies of bequest patterns to stepchildren (let alone former stepchildren) appear in the literature. (181) Sociologists have, however, examined the conditions under which emotional bonds between a stepparent and stepchild form. (182) These are most likely to arise if the child joined the stepfamily as an infant or toddler and resided within the stepparent's household. (183) Under those conditions, the resulting ties can simulate those binding parent and child. (184)

Again, whether bequests to stepchildren generally reflect the quality of their relationship with the stepparent remains unknown. Sociological studies demonstrate both the possibility and plausibility of emotional attachments after divorce, but the factors that might cause those attachments to endure--or to unravel--remain unexplored. (185) Within the area of family law, many states now recognize these relationships by granting former stepparents the right to petition for visitation rights, and cases litigating those rights are numerous; in a few instances, stepparents have even petitioned for, and been awarded, custody of stepchildren upon divorce. (186)

To be sure, these cases tell us nothing about the relative frequency with which stepparents maintain ties with former stepchildren. What the cases do reveal is that some number of stepparents are indefatigable, sticking with their stepchildren through thick and thin. Without exception, courts have refused to extend implied revocation to stepchildren or other in-laws without statutory warrant, (187) and most have declined to take judicial notice that divorce severs ties with anyone except a spouse. (188) The Commissioners' boldness is conspicuous by comparison.

Having begun with an unproven hypothesis, the Commissioners proceed to exclude extrinsic evidence that might serve to disprove it. (189) Except in limited circumstances, all of the presumptions of implied revocation created under the Code comprise conclusive presumptions. Thus, not only subjective evidence concerning the quality of the testator's relationship with a former stepchild, but also objective evidence (for example) that a testator sought visitation rights or custody of a former stepchild is inadmissible to rebut the presumption of implied revocation upon divorce. (190) The same exclusion applies to evidence concerning the quality of a testator's relationship with the former spouse, although here again intent to bequeath vel non might be clarified without undue cost. (191)

On top of this, the Code's treatment of implied revocation bristles with inconsistencies. On the one hand, the Commissioners wisely apply their rule of implied revocation for predissolution wills to will substitutes as well. (192) On the other hand, for no apparent reason, the parallel provisions of the Code covering premarital and prenatal estate plans apply exclusively to wills. (193) Another slip involves the treatment of a beneficiary named in an initial will whose status is upgraded over time. If a will includes a bequest for a person who subsequently weds the testator, the Code nevertheless grants the beneficiary an intestate share, unless the will was executed in contemplation of marriage, on the theory that the testator would probably intend to increase the bequest to reflect their changed relationship. (194) The Commissioners fail to duplicate this rule by analogy. If a will already provides for a stepchild or foster child whom the testator subsequently adopts, the child cannot claim a larger share under the Code; the Code only increases the share of a child "omitted" from a prenatal or pre-adoption will. (195)

The Code's rules of evidence exhibit similar inconsistencies. The Code allows parties to overcome the inference of intent to update a premarital will by introducing extrinsic evidence to show "that the will was made in contemplation of the testator's marriage." (196) Yet, for no apparent reason, evidence of premeditation is inadmissible with regard to either predissolution or prenatal wills: wills made in contemplation of divorce or childbirth are nevertheless subject to rules of implied revocation unless the will expressly provides otherwise. (197) The Commissioners offer no explanation for this asymmetry.

The Code also admits extrinsic evidence in the form of "the testator's statements or ... the amount of the transfer or other evidence" to show that a testator who made an inter vivos gift to a child born, or spouse wed, after the execution of a will "inten[ded] that the transfer be in lieu of a testamentary provision." (198) In other words, the Code adds a rule akin to ademption by satisfaction for bequests created by implication. Yet the Code already contains a rule of ademption by satisfaction that is readily generalizable. (199) The Commissioners have no need to reinvent this rule.

The Code fails to treat the satisfaction of express and implied bequests symmetrically. The usual doctrine of ademption by satisfaction, covering an express bequest, operates incrementally: a gift in partial satisfaction of a bequest is adeemed pro tanto. (200) Not so the parallel doctrine for satisfaction of a bequest implied by changed circumstances, which is an all-or-nothing rule. (201) In addition, the rules of evidence applicable to satisfaction of express and implied bequests differ. The only extrinsic evidence admissible to show intent to satisfy an express bequest is a "contemporaneous writing" by the testator, or an acknowledgment in writing by the beneficiary. (202) On the other hand, as just noted, intent to satisfy an implied bequest can be shown by oral declaration at any time, or simply by the scale of the inter vivos transfer. (203)

Why the asymmetry? The accompanying comments are devoid of comparative analysis. A former comment, however, called attention to the evidentiary aspect of the asymmetry (while ignoring the quantal aspect) and justified it on the ground that in connection with an implied bequest "there is no real contradiction of testamentary intent, since there is no provision in the will itself for the omitted child." (204) This argument is unpersuasive. In essence, the issue concerns the credibility vel non of extrinsic evidence, coupled with a formalized act (namely, delivery of a gift). That credibility does not hinge on whether an amendment of express language in a will or an implied bequest stands at issue. What is more, the argument proves too much: on this logic, the Code should distinguish the treatment of extrinsic evidence in connection with satisfaction of bequests and advancements of an intestate share, where again no provision in a will is implicated. Yet, the Code applies the same, stricter requirement of documentary evidence to both situations. (205)

In fact, each of the Code's alternative rules of evidence--lying on opposite sides of the common law rule--is vulnerable to criticism. As earlier observed, the common law of ademption by satisfaction admits into evidence contemporaneous declarations of intent, following the usual assumption that declarations coupled with a formalistic act are credibly performative. (206) In this light, the Code's requirement of documentary evidence to override the presumption against ademption by satisfaction appears unduly strict. But also in this light, the Code's allowance of non-contemporaneous declarations to show intent to satisfy an implied bequest appears unduly lax. (207) Once a declaration is decoupled from any formalistic act, the testator's intent that it be legally effective grows uncertain and for that reason is generally inadmissible in other contexts. (208)

CONCLUSION

This Article has argued that inheritance law should pursue an approach to testamentary obsolescence premised on the practical latitude testators have to update their own estate plans. Given the many and varied settings in which these texts come into effect, the theory developed here suggests a more complex solution to the problem of obsolescence than is ordinarily posited for its counterparts in the constitutional or legislative arenas. (209) Current law conforms to the theory in some respects, but not others. The theory, for example, supports dynamic interpretation of the wills of testators slain by beneficiaries in more situations than inheritance law presently recognizes. By contrast, the theory suggests scaling back the doctrine of implied revocation into a temporary presumption, making only provisional allowances for consequential events. Still, data are required to confirm these hypotheses.

Indeed, in every situation where theory warrants intervention to effectuate probable intent, we need data to guide our course. Some data are available today, but we must have more. Without data to inform our law, we are flying blind and cannot tell how far off target our hunches and conjectures are carrying us.

This refrain has sounded before. More than sixty years ago, a leading inheritance scholar of his era, Professor William Page, issued a like complaint:
   Even when at their very best, neither courts nor legislatures have
   made any ... systematic attempt to follow up these rules and
   principles and to see how they work in real life.... [W]hether they
   achieve the desired result in the majority of cases or whether they
   defeat it, is rarely learned by courts or legislatures, and then
   only by chance. No branch of government is charged with this duty.
   No appropriations are made to secure this result. On a priori
   theories our law is largely built up; and a priori our theories
   often remain from start to finish. (210)


If lawmaking bodies will not fill this void, who can? Is this not ideal work for law modelers--that is, the law reform organizations promulgating exemplary statutes and rules for adoption by the states? By proceeding on the basis of hard data, the Uniform Law Commission and the American Law Institute could distinguish their handiwork from that of legislative drafting committees and courts. But in fact neither organization funds empirical research for its projects. (211) Only the occasional, random academic study illuminates patches of the field.

However unfortunate this state of affairs, lawmakers (and law modelers) must soldier on regardless, catching as catch can the evidence available. As the Reporter for the Uniform Probate Code observes, "requiring a systematic empirical study before any reform can be put into place would paralyze the law-reform process." (212) The point is well taken, yet the Commissioners--perhaps seeking to put the best face on a bad situation--have staked out a position on the subject of empirical evidence that is too complaisant. The Reporter continues:
   [The concern] that there is no systematic empirical study
   supporting section[s of the Code] ... might be justified if the
   law-reform process were in the hands of amateurs who know little
   about the field or the practice.... [T]he Joint Editorial Board for
   the Uniform Probate Code ... is an organization that counts among
   its members not only leading scholars in the field but also
   nationally known estate planners of considerable insight and
   experience.... Their cumulative experience suggests that they have
   a pretty good idea of what most clients want. (213)


Truth be known, this sort of armchair empiricism holds little epistemic value. (214) Yet, by relying on their own experience where necessary as a pis aller, the Commissioners may have gotten into the habit of trusting it generally. They have followed their instincts in the face of hard data in some instances (215) and have failed to marshal related data in others. (216) This modus operandi will not recommend their products to legislative drafting committees, which can also call upon experience to gauge probable intent. One educated guess is as good as another.

At the end of the day, crafting superior rules of inheritance law will require the services not of the meta-theoretical scholar, nor of the seasoned estate planner, but of the humble pollster and statistician. Blessed are the meek, for they shall inherit the earth.

APPENDIX

We here model the concept of an error-minimizing default rule. We begin by analyzing the problem in the context of implied revocation of wills and then turn briefly to broader applications.

Stipulate that lawmakers face a binary choice between implementing default rule x or y. Default rule x effects the result most parties prefer (the majoritarian default). In connection with consequential events that tend to change intent, default rule x would create a doctrine of implied revocation. Assume that the alternative, default rule y, comprises the rule that most persons who are ignorant of what rule applies believe is in effect. In connection with consequential events that do not cause impossibility, we assume default rule y would maintain the original estate plan.

Let T be the total number of persons in the population (or sample) who are subject to a default rule, and let Nx be the number of persons who prefer outcome x; Ny prefer outcome y. Stipulate that Nx > Ny.

Assume that a randomly distributed fraction of T, which we designate as (i), is ignorant of the default rule, while the remaining fraction (1-i) know or learn of the rule. Among ignorant parties, we simplify by assuming that all share the expectation that the rule is y. This simplification follows from the surmise--unproven, and requiring empirical proof, but seemingly plausible--that doctrines of implied revocation are counter-intuitive; ignorant testators assume their wills remain in effect until they amend them. (217) As concerns other inheritance defaults, such as lapse and ademption by extinction, testators may harbor a wider range of assumptions, reducing the potential efficiency of expectation-based inheritance defaults by diluting or scattering error costs.

Assume further that a randomly distributed fraction of T, which we designate as (p), procrastinate, failing to amend their wills in response to changed circumstances, while the remaining fraction (1-p) are vigilant estate planners. Notice that (p) in turn is a function of time: as more time elapses following a consequential event, (p) should become smaller.
If x is the rule:                      If y is the rule:

(1-i)(1-p)Nx bear no cost              (1-i)(1-p)Nx bear a transaction
                                         cost
(1-i)(1-p)Ny bear a transaction cost   (1-i)(1-p)Ny bear no cost
(i)(1-p)Nx bear a transaction cost     (i)(1-p)Nx bear a transaction
                                         cost
(i)(1-p)Ny bear an error cost          (i)(1-p)Ny bear no cost
(p)Nx bear no cost                     (p)Nx bear an error cost
(p)Ny bear an error cost               (p)Ny bear no cost


We relate transaction costs and error costs with a coefficient (e), determined normatively, such that error costs = (e)transaction costs.

Under what conditions is default rule y (the error-minimizing default) more efficient than default rule x (the majoritarian default)? Where:

(1-i)(1-p)Nx + (i)(1-p)Nx + (e)(p)Nx < (1-i)(1-p)Ny + (i)(1-p)Nx + (e)(i)(1-p)Ny + (e)(p)Ny

which reduces to:

(1-i)(1-p)Nx + (e)(p)Nx < (1-i)(1-p)Ny + (e)(i)(1-p)Ny + (e)(p)Ny

Notice that as the coefficient (e) increases--that is, as we accord greater weight to error costs--the equation approaches:

(e)(p)Nx < (e)(i)(1-p)Ny + (e)(p)Ny

which we can express as:

Nx/Ny < i/p + (1-i)

Hence, as the ratio of Nx to Ny grows larger, the condition becomes increasingly difficult to meet. Meanwhile, as the fraction (p) grows smaller (which it should over time) the condition becomes increasingly easy to meet. If the fraction (p) is 1--that is, if all persons procrastinate, which they do (no matter how vigilant they are) if death follows on the heels of a consequential event--then the right side of the equation becomes 1. Because Nx > Ny, the condition is not satisfied. Therefore, assuming (e) is large, an error-minimizing default is inefficient in the short term following a consequential event.

The normative justification for making the coefficient (e) large is that error costs are negatively correlated with wealth: The poorer the testator, the more likely he or she is to bear error costs, especially because of ignorance, since information is costly--hence, the often articulated policy against making the law of wills "a mere trap and pitfall." (218) At the same time, we likely have a second, purely economic reason to discount transaction costs. Empirical evidence may well show that few testators whose intent changes following a consequential event rely on a doctrine of implied revocation to save the expense of executing a new will. Even if the doctrine would operate to carry out their intent, vigilant testators may prefer to keep on the safe side and set out their revised estate plan explicitly. (219) If all (or virtually all) vigilant testators will bear a transaction cost in any event, then only error costs are of consequence.

Assuming, nevertheless, we let (e) be a smaller number, so that transaction costs remain significant in relation to error costs, then the primary equation reduces to:

Ny/Nx < 1 - p - i + ip + ei - eip + ep/1 - p - i + ip + ep

We can readily determine that the condition may or may not be satisfied. Hold (i) constant, and let (p) vary from 0 to 1. At p = 1 (all persons procrastinate, the case where death follows hard on a consequential event), the equation becomes:

Nx/Ny < 1

Again, because we have stipulated that Nx > Ny, the condition is not satisfied. Thus, irrespective of whether transaction costs are significant in relation to error costs, an error-minimizing default is never efficient in the short term following a consequential event. On the other hand, at p = 0 (no one procrastinates, a state we may approach over time, although we may also approach a limit above 0), the equation becomes:

Nx/Ny < 1 - i + ei/ 1 - i

The condition may or may not be satisfied. As (i) approaches 0 (no ignorance either), the right side of the equation approaches 1 and, again, the condition is not satisfied. But as (i) approaches 1 (universal ignorance), the right side of the equation approaches 8, and the condition is satisfied.

Now hold (p) constant, and let (i) vary from 0 to 1. At i = 0 (no parties are ignorant), the equation becomes:

Nx/Ny < 1

Again, the condition is not satisfied. At i = 1 (all parties are ignorant), the equation becomes:

Nx/Ny < 1/p which we can alternatively express as: Ny/Nx > p

The condition may or may not be satisfied. As p approaches 1, the condition is not satisfied, but as p approaches 0 the condition is satisfied.

In connection with contracts, the problem differs somewhat. Whereas wills are ambulatory, contracts are not. Therefore, procrastination is not a factor. But if either or both contracting parties are (1) ignorant of a default rule, and (2) fall into the subset of parties who prefer a rule that they mistakenly believe applies, they may lack motivation to include in the contract an express term overriding that default rule. Furthermore, they will be prone to agree on a contract price reflecting their false assumption about the applicable rule. By setting a default rule to conform to expectations, lawmakers ensure that the price fixed under the contract is economically appropriate, given the background rules that flesh out the terms of the contract. Whether error costs outweigh transaction costs in this context, causing an error-minimizing default to be more efficient than a majoritarian default, will depend on how widespread ignorance is among contracting parties, as well as the relative weight we assign (once again) to avoiding error costs, as compared to transaction costs, in the setting of contractual exchange. Under conditions of asymmetric information, where one party is sophisticated and the other not, error-minimizing defaults also serve to prevent rent-seeking by sophisticated parties, who might prefer strategically to leave contracts incomplete in order to exploit the legal ignorance of the other party. (220)

(1.) 2 FREDERICK POLLOCK & FREDERIC W. MAITLAND, HISTORY OF ENGLISH LAW 318-20, 340 (2d ed. reissued 1968) (observing that will execution typically formed a part of the last confession); George L. Haskins, The Beginnings of Partible Inheritance in the American Colonies, 51 YALE L.J. 1280, 1289 (1942).

(2.) Statute of Uses, 1536, 27 Hen. 8, c. 10, preamble (Eng.).

(3.) The frequency of deathbed wills declined over the second half of the 19th century, Lawrence M. Friedman, Patterns of Testation in the 19th Century: A Study of Essex County (New Jersey) Wills, 8 AM. J. LEGAL HIST. 34, 37-39 (1964), a trend which may have played out over "several centuries." Id. at 39 n.17. For other empirical evidence, see for example MARVIN B. SUSSMAN ET AL., THE FAMILY AND INHERITANCE 65-69 (1970); Carol A. Engler-Bowles & Cary S. Kart, Intergenerational Relations and Testamentary Patterns: An Exploration, 23 GERONTOLOGIST 167, 168 (1983); Kristine S. Knaplund, The Evolution of Women's Rights in Inheritance, 19 HASTINGS WOMEN'S L.J. 3, 18-19 (2008).

(4.) John H. Langbein, The Nonprobate Revolution and the Future of the Law of Succession, 97 HARV. L. REV. 1108, 1135 (1984).

(5.) For still another genre of aging text, see Curtis J. Mahoney, Note, Treaties as Contracts: Textualism, Contract Theory, and the Interpretation of Treaties, 116 YALE L.J. 824 passim (2007).

(6.) See Kevin M. Stack, The Divergence of Constitutional and Statutory Interpretation, 75 U. COLO. L. REV. 1 passim (2004) (suggesting that a principled approach to the problems of statutory and constitutional obsolescence calls for their differentiation).

(7.) Although uncommon, the practice of executing codicils to wills already existed in Great Britain by the fifteenth century. 2 POLLOCK & MAITLAND, supra note 1, at 340.

(8.) For a recent discussion, see Rebecca M. Kysar, The Sun Also Rises: The Political Economy of Sunset Provisions in the Tax Code, 40 GA. L. REV. 335 passim (2006).

(9.) JOHN HART ELY, DEMOCRACY AND DISTRUST 11 (1980).

(10.) Annotation, Determination Whether Will is Absolute or Conditional, 1 A.L.R.3d 1048 (1965). Under late Roman codified law, all wills automatically became void after a period of ten years. Concern about obsolescence prompted this rule: a decade-old will was "antiquated by time ... and in reality it is impious to say that the drawing up of a plan long before death is a last will." THE THEODOSIAN CODE [section] 4.4.6, at 84 (Clyde Pharr trans., 1952) (418 AD). The rule disappeared under Justinian's code of 529 AD. THE INSTITUTES OF JUSTINIAN 300 (Thomas Collett ed. & trans., 1853).

(11.) Contingency clauses in wills appeared on occasion from the middle ages onward. 2 POLLOCK & MAITLAND, supra note 1, at 340.

(12.) Michael C. Dorf, Fallback Law, 107 COLUM. L. REV. 303 passim (2007); see, e.g., 11 U.S.C. [section] 104 (2006) (automatically adjusting dollar thresholds for inflation).

(13.) Cass Sunstein, Interpreting Statutes in the Regulatory State, 103 HARV. L. REV. 405, 494 (1989). Statutes of this sort predominated in the nineteenth century, but today they are uncommon. GUIDO CALABRESI, A COMMON LAW FOR THE AGE OF STATUTES 5 (1982).

(14.) Paul B. Sargent, Drafting of Wills and Estate Planning, 43 B.U. L. REV. 179, 190 (1963).

(15.) In re Estate of Kuralt, 15 P.3d 931, 934 (Mont. 2000); see also, e.g., In re Wilkins' Estate, 211 N.W. 652, 653-54 (Wis. 1927) (exalting freedom of testation as a "sacred right"). For additional references, see Adam J. Hirsch, Inheritance and Inconsistency, 57 OHIO ST. L.J. 1057, 1114 (1996).

(16.) Scholarly rationales for freedom of testation emphasize three considerations: its tendency to promote the production of wealth, its tendency to promote the saving of wealth, and the testator's comparative advantage to formulate a welfare-enhancing estate plan for the surviving family. For a critique, see Adam J. Hirsch & William K.S. Wang, A Qualitative Theory of the Dead Hand, 68 IND. L.J. 1, 5-14 (1992). For a rare judicial annotation, see Wogan v. Small, 11 Serg. & Rawle 141, 145 (Pa. 1824) ("[F]reedom of disposition ... is one of the greatest excitements to enterprise and industry.").

(17.) E.g., Paul Brest, The Misconceived Quest for the Original Understanding, 60 B.U. L. REV. 204, 236 (1980); Michael W. McConnell, Textualism and the Dead Hand of the Past, 66 GEO. WASH. L. REV. 1127, 1127-29 (1998); Henry P. Monaghan, Doing Originalism, 104 COLUM. L. REV. 32, 35 (2004). By labeling this argument "the dead hand problem," constitutional scholars allude, once again, to another category of potentially obsolescent text--gratuitous transfers of future interests. For a discussion, see Hirsch & Wang, supra note 16.

(18.) CALABRESI, supra note 13, at 1-7.

(19.) See the references in Adam J. Hirsch, Cognitive Jurisprudence, 76 S. CAL. L. REV. 1331, 1344 n.58 (2003). But compare Professor Bordwell:
   A surprising but more or less reassuring fact is the comparatively
   small number of changes [in inheritance statutes] that occur from
   year to year. If the statute law of wills is any criterion, our
   statute law is not the temporary, ephemeral thing that is sometimes
   supposed but for permanence and longevity compares very favorably
   with the case law.


Percy Bordwell, The Statute Law of Wills (pt. 1), 14 IOWA L. REV. 1, 3 (1928).

(20.) E.g., ANTONIN SCALIA, A MATTER OF INTERPRETATION 39-41 (1997).

(21.) The multiplication of donative texts may nevertheless complicate the updating process. Malcolm A. Moore, The Will Regenerate: From Whipping Boy to Workhorse, 7 PROB. LAW. 3, 10-12 (1981); Kent D. Schenkel, Testamentary Fragmentation and the Diminishing Role of the Will: An Argument for Revival, 41 CREIGHTON L. REV. 155, 159-62 (2008). Although the analysis in this Article focuses on wills, it is equally applicable to other non-probate estate planning texts, which either substitute for or supplement wills.

(22.) In re Jones' Estate, 60 A. 915, 922 (Pa. 1905).

(23.) Hirsch & Wang, supra note 16, at 12-13; Pamela Champine, Expertise and Instinct in the Assessment of Testamentary Capacity, 51 VILL. L. REV. 25, 48-49 (2006); see supra note 16.

(24.) For an early discussion, see Warner v. Beach, 70 Mass. (4 Gray) 162, 162-65 (1855) (rejecting the argument that "the long continued insanity of the testator" should invalidate a will made previously, when the testator was of sound mind).

(25.) Compare Professor Fellows, who asserts that the incapacitated testator's "inability to formulate subjective intent is an artificial barrier to the state's decision to impute intent," because a court can glean "what decisions the [testator] would have made if legally competent and properly advised." Mary Louise Fellows, In Search of Donative Intent, 73 IOWA L. REV. 611, 623 (1988).

(26.) Rene A. Wormser, The Doctrine of Substitution of Judgment, 9 INST. ON EST. PLAN. 15-1, 15-8 (1975).

(27.) Id. at 15-2 to 15-13.

(28.) UNIF. PROBATE CODE [section] 5-407(b)(3), 8 pt. 2 U.L.A. 386 (1998) (pre-1998 Article 5, expressly proscribing revision of wills). See also UNIF. TRUST CODE [section] 602(f) & cmt. (amended 2005), 7C U.L.A. 546 (2006) (laying down the same rule).

(29.) RESTATEMENT (THIRD) OF TRUSTS [section] 11 cmt. f (2003).

(30.) UNIF. PROBATE CODE [section] 5-411(a)(4), (7). Some testators subject to a guardianship may not lack testamentary capacity, since the two legal standards are not coextensive. The Code specifies, however, that a court should "make the least restrictive order consistent with its findings," and thus should limit guardians' powers only to the ward's "demonstrated needs." Id. [section] 4-409(b) & cmt. Therefore, if a potential ward retains testamentary capacity, the court could create a "limited guardianship" for the ward and withhold testamentary power from the guardian. See id. [section] 5-304(b)(7), (8) (distinguishing limited from unlimited guardianships).

(31.) Whereas three states have enacted the Uniform Probate Code's rule, COLO. REV. STAT. [section] 1514-411(1)(g) (2002); HAW. REV. STAT. [section] 560:5-411(a)(7) (2007); MINN. STAT. ANN. [section] 524.5-411(a)(9) (West 2007), three other jurisdictions have equivalent non-Uniform statutes. CAL. PROB. CODE [section] 2580(b)(13) (Deering 2004); NEV. REV. STAT. ANN. [section] 159.078(1)(a) (West 2005); S.D. CODIFIED LAWS [section] 29A-5-420 (1997). For a recent case applying the California statute, see Murphy v. Murphy, 78 Cal. Rptr. 3d 784 (App. Ct. 2008).

(32.) UNIF. PROBATE CODE [section] 5-411(c).

(33.) UNIF. TRUST CODE [section] 412(a). No such power existed under prior trust law apart from the more limited one, restricted to charitable trusts, created by the cy pres doctrine. RESTATEMENT (SECOND) OF TRUSTS [section] 399 (1959).

(34.) The Uniform Trust Code provision is explicitly mandatory. UNIF. TRUST CODE [section] 105(b)(4). Within guardianship law, a prospective ward can only oppose the petition, which is then judged by the court. UNIF. PROBATE CODE [section] 5-308.

(35.) Benefactors could nevertheless name their own estate planning surrogate as trustee of a discretionary trust or holder of a power of appointment to ensure flexibility upon incapacity or death, thereby effectively superseding other legal processes. See also RESTATEMENT (SECOND) OF PROP.: DONATIVE TRANSFERS [section] 34.5(3) & cmt. c, statutory note 14 (1983) (on the serviceability of a durable power of attorney for this purpose). The fact remains, however, that some fraction of benefactors will not want to entrust estate planning authority to anyone.

(36.) Or can they? The Uniform Trust Code gives a court authority to modify the terms of a trust "if, because of circumstances not anticipated by the settlor, modification ... will further the purposes of the trust," and this is ostensibly a mandatory rule. UNIF. TRUST CODE [section][section] 105(b)(4), 412(a). Yet, if the trust instrument provides that its terms should remain fixed even if circumstances change, could not one argue that the settlor has anticipated the possibility of a change of circumstance, generally if not specifically, thereby precluding recourse to modification under the plain language of the Uniform Trust Code? Similarly, under the Uniform Probate Code a provision in the will indicating that if a guardian is ever appointed he or she should not modify the will's terms even if circumstances change arguably reveals the testator's intent which, according to the Code, is supposed to guide the guardian's actions. See supra note 32 and accompanying text.

(37.) A "majoritarian" default rule minimizes transaction costs. On the public policy of default rules, see Adam J. Hirsch, Default Rules in Inheritance Law: A Problem in Search of Its Context, 73 FORDHAM L. REV. 1031, 1033-77 (2004).

(38.) See supra note 15 and accompanying text.

(39.) The traditional temporal boundary on freedom of testation is set by the Rule Against Perpetuities, extending the diktat of the will to the death of persons alive at the testator's death plus the minority of the succeeding generation. On the public policy of this rule (a large subject in itself), see Hirsch & Wang, supra note 16. The concern some benefactors may entertain that a power lodged in a third party to update their estate plans, even one monitored by courts, will turn out to thwart their intent cannot be dismissed as unreasonable. See Stephanie Strom, Donors Gone, Trusts Veer from Their Wishes, N.Y. TIMES, Sept. 29, 2007, at A1 (related discussion). Furthermore, restrictions on freedom of testation may have the perverse effect of pushing the determined benefactor to evade them in sub-optimal ways. A benefactor seeking to exercise inflexible dead hand control might thus create future interests out of trust, thereby escaping the reach of the Uniform Trust Code. Likewise, a benefactor seeking to insulate an estate plan from interference by an anticipated guardian could be driven while capable to create inter vivos future interests with a retained life estate.

(40.) Professor Wigmore's contribution has been forgotten: it goes unmentioned in modern discussions of the subject. See, e.g., Edward C. Halbach, Jr. & Lawrence W. Waggoner, The UPC's New Survivorship and Antilapse Provisions, 55 ALB. L. REV. 1091, 1091-99 (1992).

(41.) "A presumption merely shifts the duty to produce evidence ... [b]ut here it is assumed that no evidence can be produced. Therefore the solution ... is not an evidential rule of presumption, but a rule for the Distribution of Property." 9 JOHN H. WIGMORE, A TREATISE ON THE ANGLO-AMERICAN SYSTEM OF EVIDENCE IN TRIALS AT COMMON LAW [section] 2532a, at 483 (3d ed. 1940) (in the 3d ed. only).

(42.) Id. [section] 2532a, at 487.

(43.) Id. For Wigmore's proposal, see JOHN H. WIGMORE, A TREATISE ON THE ANGLO-AMERICAN SYSTEM OF EVIDENCE IN TRIALS AT COMMON LAW [section] 2532a, at 1211-13 (2d ed. Supp. 1934) (no beneficiary takes "unless a clear interval of time between the deaths is shown to have elapsed, such that the person dying later had an opportunity to make a testate disposition of the property.").

(44.) The draft is reprinted in 9 WIGMORE, supra note 41, [section] 2532a, at 484-87. Under the draft, a beneficiary did not take "unless such an interval of time between the deaths is shown to have elapsed that the survivor had a clear period of consciousness." Id. at 486.

(45.) See UNIF. SIMULTANEOUS DEATH ACT (amended 1953), 8B U.L.A. 164 (2001) (omitting the proposed provision).

(46.) UNIF. PROBATE CODE [section] 2-702 (b), (d) (amended 2008), 8 pt. 1 U.L.A. 182 (1998). A similar provision has appeared in every iteration of the Code since its debut in 1969 and now also appears in the Uniform Simultaneous Death Act, as revised in 1993. UNIF. SIMULTANEOUS DEATH ACT [section][section] 3, 6 (1993), 8B U.L.A. 149, 152 (2001). Only eleven jurisdictions have adopted a rule of this sort, however. RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 1.2 statutory note 2 (1999).

(47.) For a case litigating whether a beneficiary had satisfied or failed to satisfy an overlive requirement by five minutes, see Estate of Davies, 26 Cal. Rptr. 3d 239, 241 (Ct. App. 2005).

(48.) JOSEPH TRACHTMAN, ESTATE PLANNING 72-74 (rev. ed. 1968); Addison E. Dewey, Testators Who Die Intestate, 7 PROB. L.J. 221, 244-45 (1987) (giving examples of litigation that ensued when beneficiaries were required to survive until the estate was distributed).

(49.) JEFFREY N. PENNELL & ALAN NEWMAN, ESTATE AND TRUST PLANNING 18 (2005); JOHN R. PRICE, PRICE ON CONTEMPORARY ESTATE PLANNING [section] 4.15.4 (2d ed. 2000). Even longer overlive requirements appear sporadically in wills, e.g., In re Estate of Keller, 46 P.3d 1135, 1136 (Kan. 2002) (one-year requirement), but these are inadvisable because they delay probate and can have adverse tax consequences. PRICE, supra, [section] 4.15.4.

(50.) As early as 1932 (before Wigmore turned his attention to the problem), Ohio mandated a thirty-day overlive requirement, see Legislation, Statutory Solutions of the Problem of Survival in a Common Disaster, 50 HARV. L. REV. 344, 348 (1936), as do several Australian territories today. Wills Act, 2000, N. TERR. AUSTL. LAWS [section] 34 (2007); Succession Act, 1982, QUEENSL. PUB. ACTS [section] 33B(1) (2006); Wills Act, 1997, VICT. STAT. [section] 39 (2007). The Ohio statute was repealed in 2002, when Ohio adopted the Uniform Simultaneous Death Act. 2002 Ohio Laws file 90 (H.B. No. 242).

(51.) The two circumstances could arise in the same case, where a slayer-beneficiary commits suicide shortly after the slaying. E.g., In re Estate of Miller, 840 So.2d 703 (Miss. 2003) (construing the two rules in pari materia).

(52.) Obviously, we cannot poll victims of either common calamities or slayings, but we can present testators with hypothetical questions to discover what they anticipate their intent would be.

(53.) UNIF. PROBATE CODE [section] 2-803 (amended 2008), 8 pt. 1 U.L.A. 58 (Supp. 2008); RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 8.4 reporter's note 1 (1999). Other states bar slayers' inheritances by case law. Id. [section] 8.4 reporter's note 6.

(54.) RESTATEMENT (THIRD) OF PROP. [section] 8.4 cmt. b. For an equivalent observation, see, for example, UNIF. PROBATE CODE [section] 2-803 cmt.

(55.) Some forty states plus the federal government now prevent criminals, by analogy, from profiting out of the publicity generated by their crimes. Orly Nosrati, Note, Son of Sam Laws: Killing Free Speech or Promoting Killer Profits?, 20 WHITTIER L. REV. 949, 953 nn.48-49 (1999).

(56.) Bennett v. Allstate Ins. Co., 722 A.2d 115, 117-18 (N.J. Super. Ct. App. Div. 1998); Riggs v. Palmer, 22 N.E. 188, 192 (N.Y. 1889) (Gray, J., dissenting); In re Wilkins' Estate, 211 N.W. 652, 656 (Wis. 1927). For academic discussions, see RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW [section] 18.2 (7th ed. 2007); Mary Louise Fellows, The Slayer Rule: Not Solely a Matter of Equity, 71 IOWA L. REV. 489, 504-06 (1986); William M. McGovern, Jr., Homicide and Succession to Property, 68 MICH. L. REV. 65, 71-72 (1969); Jeffrey G. Sherman, Mercy Killing and the Right to Inherit, 61 U. CIN. L. REV. 803, 860-63 (1993).

(57.) In two jurisdictions, however, the slayer statute operates as a default rule. LA. CIV. CODE ANN. arts. 941, 943, 945 (2000) (allowing the slayer to inherit if he or she "proves reconciliation with or forgiveness by the decedent"); WIS. STAT. ANN. [section] 854.14(6) (West 2002) (allowing a decedent to override the slayer statute by specific reference in a will, and alternatively allowing the court to override the statute in deference to "the decedent's wishes").

(58.) But see Sherman, supra note 56, at 856-76 (proposing and defending exceptions from the operation of slayer statutes for assisted suicide and mercy killings).

(59.) UNIF. PROBATE CODE [section] 2-803(b).

(60.) E.g., Ford v. Ford, 512 A.2d 389, 398-99 (Md. 1986) ("[T]he slayer's rule is simply not applicable when the killer was not criminally responsible at the time he committed the homicide.... [T]he maxims prompting the rule--no one shall be permitted to profit by his own fraud ... [are] inappropriate when a person is criminally insane."). For earlier cases, see id. at 400-04; Michael G. Walsh, Annotation, Homicide as Precluding Taking Under Will or By Intestacy, 25 A.L.R.4th 787, 863-68 (1983). Slayer statutes bar inheritance by insane slayers in only two states. IND. CODE [section] 29-1-2-12.1(a) (2000); OHIO REV. CODE ANN. [section] 2105.19(A), (C) (LexisNexis 2002).

(61.) Sherman, supra note 56, at 848 n.210; Walsh, supra note 60, at 820-21.

(62.) WAYNE R. LAFAVE, CRIMINAL LAW [section] 10.4(b) (4th ed. 2003).

(63.) Direct inquiry is again impossible, but we can assess the attitudes of testators by presenting them with hypotheticals. Query, incidentally, whether testators polled ex ante would excuse self-defense if they themselves were mentally ill when they assaulted the beneficiary. Refinements may be necessary.

(64.) In eight states, however, slayer laws do apply to unintentional-but-criminal killings. Sherman, supra note 56, at 848 n.213. For academic commentary, compare POSNER, supra note 56, [section] 18.2, at 544; Fellows, supra note 56, at 496-99; McGovern, supra note 56, at 92-93.

(65.) LAFAVE, supra note 62, [section] 15.4(a), at 794-95.

(66.) Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 19 (1991).

(67.) Such conduct often constitutes involuntary manslaughter. LAFAVE, supra note 62, [section] 14.3.

(68.) See infra note 149 (Louisiana and Oregon).

(69.) The case of the heiress Sunny von Bulow is illustrative: Although acquitted of the criminal charge, her husband was alleged to have attempted to murder her by insulin injection in 1980. She survived but fell into a persistent coma, remaining in that state until her death, twenty-eight years later. Enid Nemy, Sunny von Bulow, Whose Near Death Started a Society Drama, Dies at 76, N.Y. TIMES, Dec. 7, 2008, at A42.

(70.) For a fuller statement of this rhetorical argument, see Hirsch, supra note 37, at 1042-44. See also supra note 16.

(71.) POSNER, supra note 56, [section] 4.1, at 96, [section] 18.2, at 544. For an early recognition, see HENRY SUGDEN, AN ESSAY ON THE LAW OF WILLS 221 app. (London, S. Sweet 1837).

(72.) Randy L. Buckner & Daniel C. Carroll, Self-Projection and the Brain, 11 TRENDS COGNITIVE SCI. 49 passim (2007); Daniel T. Gilbert & Timothy D. Wilson, Prospection: Experiencing the Future, 317 SCI. 1351 passim (2007); see also George Loewenstein et al., Projection Bias in Predicting Future Utility, 118 Q.J. ECON. 1209 passim (2003) (elaborating a related phenomenon). For an early recognition, see Jeremy Bentham, General View of a Complete Code of Laws, in 3 THE WORKS OF JEREMY BENTHAM 155, 190-91 (John Bowring ed., 1962) (1802).

(73.) SCOTT PLOUS, THE PSYCHOLOGY OF JUDGMENT AND DECISION MAKING 125 (1993).

(74.) THOMAS L. SHAFFER ET AL., THE PLANNING AND DRAFTING OF WILLS AND TRUSTS 14-15 (5th ed. 2007); THOMAS L. SHAFFER, DEATH, PROPERTY, AND LAWYERS 123-24 (1970); Sargent, supra note 14, at 191-92.

(75.) See supra notes 48-49.

(76.) See supra note 35.

(77.) For British data on the frequency of homemade wills, which are statistically correlated with poverty, see JANET FINCH ET AL., WILLS, INHERITANCE, AND FAMILIES 45-46 (1996).

(78.) For defenses of this norm, which Professor Fellows labels "equal planning under the law," see Fellows, supra note 25, at 613; Hirsch, supra note 37, at 1051-52; see also id. at 1044-61 (related discussion).

(79.) WILLIAM BLACKSTONE, 1 COMMENTARIES *91.

(80.) An estate plan might have been impossible to implement even when it was executed, but that raises a problem strictly of mistake, not obsolescence. We may also note an analogy to the problem of postmortem impossibility, raised in connection with future interests. Here, the cy pres doctrine applies, traditionally an intent-focused rule once again. 4A AUSTIN WAKEMAN SCOTT, THE LAW OF TRUSTS [section] 399, at 476 (William F. Fratcher ed., 4th ed. 1989).

(81.) See supra notes 71-78 and accompanying text. Since testators can revise an impossible estate plan, disparities of intent effectuation here arise not strictly out of better or worse planning, but out of greater or less diligence. Yet estate planners will also anticipate these circumstances as a matter of course and provide for them via contingency clauses--hence, in practice, opening a divide between well drafted estate plans that require no diligence and poorly drafted ones that do. As further discussed below, the expectations of the parties can affect the efficiency of an inheritance default, if ignorant testators generally share the same assumption about what the default rule is. See infra notes 114-18 and accompanying text & Appendix. Whether that is true as concerns lost property or predeceasing beneficiaries is uncertain and merits investigation, although there is no prima facie reason to anticipate a convergence of assumptions in either regard.

(82.) As one court opined:
   The process of making testamentary bequests by a will is not akin
   to the donning and doffing of a suit of clothing.... "In no other
   field of law should more care be taken over a decision to lessen
   requirements than in the field of Wills where the introduction of
   fraud must be guarded against to a greater extent because ... the
   lips of the main 'witness' (the testator) are sealed."


In re Estate of Weston, 833 A.2d 490, 492 (D.C. 2003) (quoting In re Lee's Estate, 80 F. Supp. 293, 294 (D.D.C. 1948)). For a further analysis of the justifications for will execution formalities, see Hirsch, supra note 15, at 1060-69. Declarations made by the testator to an attorney or other fiduciary are safer from fraud but still provide no assurance of finality.

(83.) For an early observation, see Humphreys v. Humphreys, (1789) 30 Eng. Rep. 85, 85 (Ch.) (Thurlow, C.) ("[D]iscussing what were the particular motives and intention of the testator in each case [of bequests of absent property] ... would be productive of endless uncertainty and confusion.").

(84.) We have some experience with a liberal rule of evidence in connection with patent and latent ambiguities, since courts nowadays do routinely look beyond the four walls of a will to clarify its meaning. Doing so has not occasioned any conspicuous mischief. The problems, though, are not identical: In the case of linguistic ambiguity, we know the testator meant something by the words he or she used; in the case of contingencies missing from the will, we must infer what the testator would have intended if he or she had considered the contingency, and that is a more problematic exercise. For an argument that courts should admit some forms of extrinsic evidence to determine intent in connection with bequests to predeceasing beneficiaries, see Susan F. French, Antilapse Statutes Are Blunt Instruments: A Blueprint for Reform, 37 HASTINGS L.J. 335 passim (1985).

(85.) RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 5.5 statutory note (1999).

(86.) In re Connolly's Estate, 222 N.W.2d 885, 893 (Wis. 1974); Hirsch, supra note 15, at 1129.

(87.) FINCH ET AL., supra note 77, at 138-43 (British data); Allison Dunham, The Method, Process and Frequency of Wealth Transmission at Death, 30 U. CHI. L. REV. 241, 282-84 (1963). Professor Browder considered this option "perfectly natural, particularly when the [predeceasing beneficiary's] issue are likely to be minors," although his data contained few examples of this preference. Olin L. Browder, Jr., Recent Patterns of Testate Succession in the United States and England, 67 MICH. L. REV. 1303, 1322-27 (1969) (quotation at 1326). Cf. Halbach & Waggoner, supra note 40, at 1101-02 (defending the traditional rule). For the one state statute deviating from the rule that only progeny can serve as substitutes, see MD. CODE ANN., EST. & TRUSTS [section] 4-403 (LexisNexis 2001). In several Canadian provinces, the beneficiary's surviving spouse can take under antilapse statutes in certain circumstances. See, e.g., Wills Act, R.S.B.C. [section] 29 (1996); Wills Act, NFLD. R.S. [section] 18(1) (1990); Succession Law Reform Act, R.S.O. [section] 31 (1990).

(88.) Whether typical testators intend a different substitute for predeceasing relatives and nonrelatives needs inquiry. The drafters of the first British antilapse statute of 1837 who ushered in this structural distinction, still found in modern statutes, did not justify it as intent-effectuating--that appears to have been a subsequent rationalization. See supra note 86 and accompanying text. Rather, the British drafters were "not disposed to recommend alterations of the present laws further than cases of frequent occurrence and great hardship appear to render necessary; and we therefore confine our recommendation of an amendment in this respect to the cases of ... children and grandchildren of the testator." SUGDEN, supra note 71, at 222 app. For commentary questioning the restriction of the antilapse concept, see JESSE DUKEMINIER ET AL., WILLS, TRUSTS, AND ESTATES 393 (7th ed. 2005); French, supra note 84, at 365-66.

(89.) See supra note 85.

(90.) Under the Code, the antilapse statute covers bequests to stepchildren (but not other affines), along with close blood relatives. UNIF. PROBATE CODE [section] 2-603(b) (amended 2008), 8 pt. 1 U.L.A. 49 (Supp. 2008). The accompanying comment asserts that the Code establishes a rule of "presumed intention" but cites no evidence, although some pertinent data does exist. Id. [section] 2-603 cmt.; supra note 87. At the same time, the Code applies a different antilapse rule (applicable to all beneficiaries) with respect to future interests, including living trusts, id. [section] 2-707(b), (d), even though these are functionally equivalent to wills. Regarding stepchildren, the Commissioners apparently infer that if a testator bequeaths to one, this fact implies they have formed a bond comparable to a consanguineous tie. See Halbach & Waggoner, supra note 40, at 1121-23 (offering an example). But why then does the Code not extend this inference to bequests to step-grandchildren, or to step-siblings? The accompanying comment offers no rationale, although it notes the statutory distinction. UNIF. PROBATE CODE [section] 2-603 cmt.; cf. CAL. PROB. CODE [section] 21110(c) (Deering 2004) (extending the antilapse statute to all kin of a spouse). Query also whether the relative size of a bequest is a significant variable. Arguably, when a bequest to a non-relative comprises a substantial fraction of the estate, its very scale implies that the beneficiary comprises virtual family. French, supra note 84, at 364-67. For a variety of suggestions for statutory reform of antilapse doctrine, see id. at 363-73.

(91.) DUKEMINIER ET AL., supra note 88, at 405-06.

(92.) In re Estate of Fox, 431 A.2d 1008, 1010 (Pa. 1981) (quoting Hoke v. Herman, 21 Pa. 301, 305 (1853)). "[I]ntention has nothing to do with the matter." In re Wright's Will, 165 N.E.2d 561, 562 (N.Y. 1960). For additional references, see Hirsch, supra note 15, at 1128-29 & nn.215, 218.

(93.) On the potential relevance of this variable, see Hirsch, supra note 15, at 1134 n.229.

(94.) In Kentucky, bequests to persons who would comprise the testator's heirs do not adeem, whereas bequests to other persons do adeem. KY. REV. STAT. ANN. [section] 394.360 (LexisNexis 1999).

(95.) Would the typical testator distinguish ex ante the testamentary consequences of voluntary alienation of property from involuntary or accidental alienation, for example? On this distinction, see WILLIAM M. MCGOVERN & SHELDON F. KURTZ, WILLS, TRUSTS AND ESTATES [section] 8.1, at 318-19 (3d ed. 2001); Hirsch, supra note 15, at 1130-31 & n.222.

(96.) Under the Uniform Probate Code, beneficiaries receive any property that the testator acquired "as a replacement," for specifically bequeathed property that is lost or disposed of--thus construing specific bequests as, in effect, generic bequests. UNIF. PROBATE CODE [section] 2-606(a)(5), 8 pt. 1 U.L.A. 51 (Supp. 2008).

(97.) Id. [section] 2-606. For a further discussion, see Hirsch, supra note 15, at 1130-35. A related problem arises where the satisfaction of creditors' claims against the testator (which take priority over bequests) leaves insufficient funds in the estate to carry out the estate plan in its entirety. In that event, under the doctrine of abatement, debts are satisfied first by reducing residuary bequests; then, if necessary, by reducing general bequests (of amounts, typically money) pro rata; then, if necessary, by reducing specific bequests (of items) pro rata. UNIF. PROBATE CODE [section] 3-902(a). The Commissioners submit, again without evidence, that this traditional order of abatement "may be regarded as approximating what testators generally want." Id. [section] 3-902 cmt. Yet, when instead reducing bequests to satisfy tax claims or the spouse's elective share, the Commissioners follow a different order of abatement, taking a proportional share from all beneficiaries. Id. [section][section] 2-207(b), 3-9A-104(1).

(98.) The Uniform Probate Code is exceptional, opening the door even to ex post declaratory evidence: beneficiaries receive a pecuniary devise equal in value to specifically bequeathed property that is lost or disposed of if "ademption would be inconsistent with the testator's manifested plan of distribution or [if] at the time the will was made, the date of disposition or otherwise [i.e., at a later time?] the testator did not intend ademption of the devise." UNIF. PROBATE CODE [section] 2-606(a)(6). This provision has not been adopted in a single jurisdiction, however. A prior version of the provision made a presumption against ademption "unless the facts and circumstances indicated that ademption of the devise was intended ... or ... is consistent with the testator's manifested plan of distribution"--again opening the door to extrinsic evidence. UNIF. PROBATE CODE [section] 2-606(a)(6), 8 pt. 1 U.L.A. 176 (1998) (pre-1997 Article 2) (adopted in 4 jurisdictions: Colo., Mich., Mont. and Utah). See also UNIF. PROBATE CODE [section] 2-601 ("In the absence of a finding of a contrary intention, the rules of construction [for lapse and ademption] control the construction of a will.") (adopted in 11 jurisdictions). This provision could be read to allow consideration of extrinsic evidence; but because it is supposed "to align the statutory rules of construction ... with those established at common law," id. [section] 2-601 cmt., which currently bars extrinsic evidence, 6 WILLIAM J. BOWE & DOUGLAS H. PARKER, PAGE ON THE LAW OF WILLS [section] 50.1, at 71-72, [section][section] 54.14-.15 (rev. ed. 2005); Hirsch, supra note 15, at 1128-29, the provision will admit of multiple interpretations. This provision also does not cover the doctrine of abatement, but the separate section covering that subject allows a court to deviate from the statutory order of abatement if it would defeat "the express or implied purpose of [a] devise." UNIF. PROBATE CODE [section] 3-902(a) (emphasis added). Although the language here is again frustratingly vague, it could be read to allow a court to consider non-declaratory extrinsic evidence. See id. [section] 3-902 cmt. By comparison, the Code makes no allowance for consideration of extrinsic evidence in connection with abatement to satisfy tax claims, see id. [section] 3-9A-104(1), nor in connection with abatement to satisfy the spouse's elective share, unless a seemingly superfluous adverb connotes greater latitude: "liability ... is equitably apportioned among the [beneficiaries] ... in proportion to the value of their interests." Id. [section] 2-207(b) (emphasis added); see also id. [section] 2-207 cmt. (failing to address the meaning of this phrase).

(99.) RESTATEMENT (SECOND) OF PROP.: DONATIVE TRANSFERS [section] 5.2 (1983).

(100.) Id. [section] 5.2 cmts. c-h & reporter's note 4. Nevertheless, ex post declarations are inadmissible:
   Where impossibility is offered as an excuse for non-performance,
   the construction process required under the stated rule is an
   appropriate inquiry, not as to what the transferor actually
   intended, but as to what his intent would have been had he known of
   or anticipated the impossibility at the time he imposed the
   restraint.


Id. [section] 5.2 cmt. c.

(101.) Id. [section] 5.2 cmt. c.

(102.) See Aten v. Tobias, 220 P. 196, 202 (Kan. 1923) ("[T]he testator must necessarily have thought of the consequences ... unless he bestirred himself to alter the testamentary disposition already made of his personalty. But the testator was content to let it stand as made, and the courts may not meddle with it.").

(103.) Brush v. Wilkins, 4 Johns. Ch. 506, 512 (N.Y. Ch. 1820) (Kent, C.) (explaining the evolution of the doctrine of implied revocation, discussed below, in the English courts).

(104.) New Jersey is exceptional in this regard; its "doctrine of probable intent" appears to permit this sort of inquiry. N.J. STAT. ANN. [section] 3B:3-33.1 (West 2007); In re Estate of Tateo, 768 A.2d 243, 246-48 (N.J. Super. Ct. App. Div. 2001). For foreign legislation recreating this doctrine, see Wills Act, 1968, AUSTL. CAP. TERR. LAWS [section] 12A(2) (1991).

(105.) In re Grave's Estate, 457 P.2d 71, 77 (Kan. 1969).

(106.) Id. For a recent reiteration, see In re Estate of Nash, 164 S.W.3d 856, 861 (Tex. App. 2005), aff'd 220 S.W.3d 914 (Tex. 2007). The Commissioners agree that except in enumerated situations, "a change of circumstances does not revoke a will or any part of it." UNIF. PROBATE CODE [section] 2-508 (amended 2008), 8 pt. 1 U.L.A. 154 (1998); see also id. [section] 2-804(f). Although the Third Restatement includes a cause of action to reform a will on the ground of mistake, changes of circumstance are excluded from the scope of the doctrine. RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 12.1 cmt. h & illus. 3 (1999).

(107.) In the language of one court:
   Were we to ... inquire into the circumstances of a man's estate, at
   the time of making his will, and of his death, we should find
   innumerable instances in which the alteration has been so great,
   that a prudent man would have altered his will, and part of his
   family has been reduced to great distress by his not doing it. But
   it is better that individuals should be distressed, than the
   freedom of disposition by last will invaded.... Once [we] establish
   the judicial habit of examining the situation of a man's fortune or
   family, and revoking his will [by inference] ... and no man's will
   is safe.


Wogan v. Small, 11 Serg. & Rawle 141, 144 (Pa. 1824). "Were it otherwise, the statute of wills would be virtually abolished." Sneed v. Ewing, 28 Ky. 460, 473-74 (1831); Hertrais v. Moore, 88 N.E.2d 909, 912 (Mass. 1949) ("It would be a serious matter to invalidate a will because of a supposed change of intention."); Hoitt v. Hoitt, 3 A. 604, 614, 616 (N.H. 1886) (such a rule would "inevitably invite litigation and 'produce infinite uncertainty and delay in the settlement of estates.'"); Brush, 4 Johns. Ch. at 519 (Kent, C.) ("It might be dangerous and lead to loose speculations ... running the hazard of substituting [the court's] will for that of the testator."); White v. Barford, (1815) 105 Eng. Rep. 739, 739 (K.B.) ("But where are we to stop? Is the rule to vary with every change which constitutes a new situation ...?"); LEWIS M. SIMES & PAUL E. BASYE, PROBLEMS IN PROBATE LAW INCLUDING A MODEL PROBATE CODE [section] 53 cmt. (1946) ("Such a doctrine introduces an undesirable element of uncertainty into ... the validity of a ... will.").

(108.) For several additional triggering mechanisms peculiar to particular states, see infra note 149.

(109.) Although today these doctrines are ordinarily statutory, their roots lie in English common and ecclesiastical law, with still deeper roots in Roman law. W.A. Graunke & J.H. Beuscher, The Doctrine of Implied Revocation of Wills by Reason of Change in Domestic Relations of the Testator, 5 WIS. L. REV. 387, 387-94 (1930). For modern discussions, see 2 BOWE & PARKER, supra note 98, [section][section] 21.86.111; McGOVERN & KURTZ, supra note 95, [section][section] 3.5-.6, 5.4. Statutes cover predissolution wills in 48 states, premarital wills in 34 states, and prenatal wills in 47 states. Although implied revocation remains a common law doctrine in Mississippi, Hinders v. Hinders, 828 So.2d 1235, 1241 (Miss. 2002), in most states today the statutes operate exclusively.

(110.) DUKEMINIER ET AL., supra note 88, at 269; Graunke & Beuscher, supra note 109, at 387.

(111.) 4 JAMES KENT, COMMENTARIES ON AMERICAN LAW *521 (Oliver Wendell Holmes ed., 12th ed. 1896) (1826-30); see also Brush, 4 Johns. Ch. at 511, 518 (Kent, C.) (putting this theory into practice).

(112.) Coughlin v. Bd. of Admin., Pub. Employees' Ret. Sys., 199 Cal. Rptr. 286, 287 (Ct. App. 1984). The same rationale appears explicitly in scores of cases, e.g., In re Estate of Rodriguez, 160 P.3d 679, 686 (Ariz. Ct. App. 2007), and dipping back to the nineteenth century, e.g., Baacke v. Baacke, 69 N.W. 303, 304 (Neb. 1896). The Commissioners likewise defend the doctrines as serving to rework an estate plan in the manner "the testator would want ... if he or she had thought about the relationship of his or her old will to the new situation." UNIF. PROBATE CODE [section] 2-301 cmt., 8 pt.1 U.L.A. 133 (1998). At the same time, a few courts rationalize implied revocations as implementing social norms: "Strong public policy upholding the institution of marriage prohibits a man from inheriting from a woman whom he ... divorced." In re Estate of Pekol, 499 N.E.2d 88, 90 (Ill. App. Ct. 1986); In re Estate of Forrest, 706 N.E.2d 1043, 1046 (Ill. App. Ct. 1999) (citing to Pekol). For an argument against default rules based on social norms, in contradistinction to probable intent, see Hirsch, supra note 37, at 1046-52; see also id. 1042-46 (on the limited potential efficiency of social defaults).

(113.) For a rare judicial recognition, see Luff v. Luff, 359 F.2d 235, 240 (D.C. Cir. 1966) (Leventhal, J., dissenting) ("Doubtless many, perhaps most, divorced men desire to disinherit their wives.... What we are considering, however, is the probable intent of those divorced men who do not destroy their wills.").

(114.) "The statutes ... anticipate that [testators] themselves will often fail to ... revoke, not out of conscious intent, but simply from a lack of attentiveness." Coughlin, 199 Cal. Rptr. at 287-88. "Legislatures ... acknowledged the lax tendencies of the public concerning the making or reviewing of wills ... [and] took a parens patriae approach to protect our citizens from the repercussions of their probable inaction." In re Estate of Forrest, 706 N.E.2d at 1045 (paraphrasing In re Estate of Knospe, 626 N.Y.S.2d 701, 703 (Sur. Ct. 1995)); In re Estate of Jackson, 194 P.3d 1269, 1274 (Okla. 2008) (the statute "is an assurance ... [against] unintentional[] omi[ssions] from a will.").

(115.) The instant analysis jumps off from an earlier inquiry of mine into the impact of information costs on default rule theory. Hirsch, supra note 37, at 1041-42.

(116.) Empirical studies have found widespread ignorance of the rules of intestacy, by analogy:
   Although there is a demonstrated interest within the general
   population about the distribution of property at death, ... an
   overwhelming majority of the citizenry are not aware of the present
   pattern of distribution provided under the intestate succession
   statutes. Therefore, they do not rely intentionally on those
   statutes to dispose of their property. Thirty percent of the
   respondents admitted they did not know the identity of their
   potential heirs and 64 percent of those who thought they knew were
   in error.


Mary Louise Fellows et al., An Empirical Study of the Illinois Statutory Estate Plan, 1976 U. ILL. L.F. 717, 723 (emphasis added) (footnote omitted) [hereinafter Fellows et al., Empirical Study]. For comparable findings, see Mary Louise Fellows et al., Public Attitudes About Property Distribution at Death and Intestate Succession Laws in the United States, 1978 AM. B. FOUND. RES. J. 319, 340 [hereinafter Fellows et al., Public Attitudes] (55.4% in error); see also id. at 340 n.73 (observing that no respondent was aware of "subtleties in the statute"); Joel R. Glucksman, Intestate Succession in New Jersey: Does It Conform to Popular Expectations?, 12 COLUM. J.L. & SOC. PROBS. 253, 262-66 (1976).

(117.) The point has not entirely escaped notice by lawmakers. One judge criticized the invocation of a doctrine of implied revocation to override a will: "It is fair to observe that people tend to do the things they really want to do, and that inaction generally signifies contentment or at worst indecision," Luff, 359 F.2d at 240 (Leventhal, J., dissenting), while also noting the problem of legal ignorance:
   A man who intends to disinherit a divorced wife is more likely to
   speak up to his counsel at once, and have it taken care of. A man
   who has decided not to disinherit his divorced wife is less likely
   to bring the matter up even assuming he is aware of the
   little-known statutory technique of republication of a will.


Id. at 242. Another court defended its limited version of the doctrine of implied revocation, see infra text at notes 145-46, as "[meeting] the expectations of those involved ... in most cases." Aetna Life Ins. Co. v. Wadsworth, 689 P.2d 46, 52 (Wash. 1984); see also In re Estate of Blanchard, 218 N.W.2d 37, 39 (Mich. 1974) (observing that in the event of divorce followed by remarriage a testator "would see no necessity to republish" a predissolution will). Also showing awareness of the problem, legislators in California have mandated that petitions for, and judgments of, dissolution of marriage must each contain a warning notice: "Dissolution ... of your marriage may automatically cancel your spouse's rights under your will.... If these are not the results that you want, you must change your will ... to reflect your actual wishes." CAL. FAM. CODE [section] 2024 (Deering 2006). See also infra note 162. But cf. Johnston v. Laird, 52 P.2d 1219, 1223 (Wyo. 1935) ("Would it not be more logical to believe that after his ... divorce, ... if he still desired this will to control the disposition of his property, he would have published and redeclared the same, so that his intention could not be questioned?") (quoting In re Gilmore's Estate, 260 N.Y.S. 761 (Sur. Ct. 1932)).

(118.) For a taxonomy of default rules and their functions, see Alan Schwartz, The Default Rule Paradigm and the Limits of Contract Law, 3 S. CAL. INTERDISC. L.J. 389, 390-91 (1993).

(119.) E.g., Warner v. Beach, 70 Mass. (4 Gray) 162, 162 (1855) (including a contingency for whether the beneficiary in utero was a son or a daughter).

(120.) Terror management theory leads persons to underestimate this risk. For a discussion and application of terror management theory in a different context, see Hirsch, supra note 37, at 1049-51; see also supra note 73 and accompanying text.

(121.) The common law failed to distinguish between beneficiaries born after a will was executed and those who were also born posthumously, applying the same doctrine of implied revision to each. C.H. SHERRIN & R.F.D. BARLOW, WILLIAMS' LAW RELATING TO WILLS 528-30 (4th ed. 1974); Belton v. Summer, 12 So. 371, 373 (Fla. 1893); Warner, 70 Mass. (4 Gray) at 163-64 (dictum). Whether modern statutes reserving a share for children under prenatal wills protect posthumously born children depends on whether an additional statute treats children in gestation at the testator's death as living at that time. Under the Uniform Probate Code, such a provision used to exist, but in 2008 it was amended to cover only "intestate succession, homestead allowance, and exempt property." Compare UNIF. PROBATE CODE [section] 2-104(a)(2) (amended 2008), available at http://www.law.upenn.edu/bll/ archives/ulc/upc/2008amends.htm, with id. [section] 2-108, 8 pt. 1 U.L.A. 87 (pre-2008 Article 2).

(122.) Although no empirical study addresses this issue, it has arisen indirectly, and anecdotally, in connection with prenuptial gifts. In New York, if either party voluntarily breaks an engagement, the law creates a "strong presumption ... that any gifts made during an engagement period are given solely in consideration of marriage, and are recoverable if the marriage does not materialize." Friedman v. Geller, 368 N.Y.S.2d 980, 981 (Civ. Ct. 1975). Nevertheless, when a donor died prior to marriage, one court denied recovery to the donor's family on the ground of implicit intent: "I firmly believe that had [the donor] thought of these consequences he would have intended that in the event of his untimely death [the donee] should keep the ring as a symbol of his love and affection." Cohen v. Bayside Fed. Sav. & Loan Ass'n, 309 N.Y.S.2d 980, 983 (Sur. Ct. 1970).

(123.) "[I]t is improbable that at the time of the [engagement] gift either [party] gave a thought to the consequences that would arise in the event of the death of one of the parties." Cohen, 309 N.Y.S.2d at 983. See also supra notes 73, 120 and accompanying text.

(124.) DUKEMINIER ET AL., supra note 88, at 413. See supra text at note 91.

(125.) In such event, to recall an old common-law maxim, "the testator ... becomes his own executor." Beck v. McGillis, 9 Barb. 35, 57 (N.Y. 1850).

(126.) On the common law of ademption by satisfaction, see SHERRIN & BARLOW, supra note 121, at 242-53; Barney Barstow, Ademption by Satisfaction, 6 WIS. L. REV. 217 passim (1931).

(127.) SHERRIN & BARLOW, supra note 121, at 243, 245-46; Barstow, supra note 126, at 218-22. The presumption for children only applies to substantial and extraordinary gifts, thus excluding periodic gifts for support. SHERRIN & BARLOW, supra note 121, at 243-44; Barstow, supra note 126, at 121-22.

(128.) UNIF. PROBATE CODE [section] 2-609 (amended 2008), 8 pt. 1 U.L.A. 179 (1998). Half the states track the Uniform Probate Code on these points. RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 5.4 statutory note (1999).

(129.) Academic commentators have not assayed this anomaly. For a judicial recognition of it, but failing to address its basis in policy, see Rhodes v. Kebke, 167 S.W.2d 345, 347-48 (Tenn. 1943).

(130.) RESTATEMENT (SECOND) OF PROP.: DONATIVE TRANSFERS [section] 31.2 & cmt. a, illus. 1-4 (1983). Similarly, contemporaneous declarations are admissible to clarify the legal ramifications of revocation by act. Hirsch, supra note 15, at 1089-96.

(131.) UNIF. PROBATE CODE [section] 2-109 cmt.; see also RESTATEMENT (THIRD) OF PROP. [section] 5.4 cmt. b (chaired by the same reporter, and repeating this claim).

(132.) Lawrence H. Averill, Jr., An Eclectic History and Analysis of the 1990 Uniform Probate Code, 55 ALB. L. REV. 891, 915-16 (1992) (questioning whether the presumption is appropriate for large gifts); see Walters v. Stewart, 588 S.E.2d 248, 249 (Ga. Ct. App. 2003), rev'd sub nom. Stewart v. Walters, 602 S.E.2d 642 (Ga. 2004) (finding evidence that a $50,000 gift to a child was intended to satisfy a bequest, although the testator failed to say so in a contemporaneous writing); see also Mary L. Fellows, Concealing Legislative Reform in the Common-Law Tradition: The Advancements Doctrine and the Uniform Probate Code, 37 VAND. L. REV. 671, 704-06 (1984) (criticizing the analogous Code provision covering gifts by intestate benefactors). Empirical studies show that most testators bequeath equal amounts to children irrespective of relative need, a tropism that one study attributes to a desire to signal equality of affection. B. Douglas Bernheim & Sergei Severinov, Bequests as Signals: An Explanation for the Equal Division Puzzle, 111 J. POL. ECON. 733 passim (2003). Although other studies find that inter vivos gifts (by comparison) are distributed unequally, id. at 734, these studies fail to explore whether donors prefer that gifts satisfy bequests. The instant study posits that parents make unequal gifts because, unlike bequests, they can keep them secret from non-recipient children and thereby maintain the image of equality of affection. Id. at 735, 752-53. Yet large gifts are probably harder to hide from siblings than small ones, suggesting that parents should prefer to equalize total sizable transfers during life and at death, assuming this model is correct.

(133.) UNIF. PROBATE CODE [section] 2-109 cmt.

(134.) THE HAMMURABI CODE [section] 165, at 32 (Chilperic Edwards ed. 1921) (c. 2084-81 BC).

(135.) Id. [section] 166, at 32-33.

(136.) UNIF. PROBATE CODE [section] 2-606(b), (d). Temporary presumptions also arise in other fields of law including, for example, bankruptcy law. 11 U.S.C. [section] 524(m)(1) (2006).

(137.) For a criticism of this provision, see Hirsch, supra note 15, at 1130-32.

(138.) See In re Estate of McQuay, 335 N.E.2d 746, 750 (Ohio Ct. App. 1975) (noting that illness can interfere with estate planning); Curley v. Giltrop, 496 N.E.2d 224, 225 (N.Y. 1986) (same).

(139.) See Gay C. Kitson & Leslie A. Morgan, The Multiple Consequences of Divorce: A Decade Review, 52 J. MARRIAGE & FAM. 913, 913-15 (1990) (linking divorce with psychological or physical morbidity).

(140.) The high court of Tennessee observed:
   A divorce does not come as an earthquake, without warning, to
   suddenly break apart a marriage.... The parties have months in
   which to contemplate and consider all the ramifications of a
   divorce.... In such a situation the number of days after the
   divorce is granted is far less significant and should not give rise
   to a presumption [of revocation].


Bowers v. Bowers, 637 S.W.2d 456, 458 (Tenn. 1982); see also Marilyn Coleman et al., Divorce and Postdivorce Relationships, in THE CAMBRIDGE HANDBOOK OF PERSONAL RELATIONSHIPS 157, 15859 (Anita L. Vangelisti & Daniel Perlman eds. 2006) (surveying the literature on sociological processes preceding and leading toward divorce). But does not ripeness once again come into play here, delaying revision of a will when divorce is contemplated but not yet certain to occur?

(141.) HENRY SWINBURNE, A BRIEF TREATISE OF TESTAMENTS AND LAST WILLS 269 (photo. reprint 1978) (1590) (emphasis added) (footnotes omitted).

(142.) In re Hall's Estate, 119 N.W. 219, 222 (Minn. 1909).

(143.) Brown v. Grand Lodge A.O.U.W., 57 A. 176, 177-78 (Pa. 1904).

(144.) See Card v. Alexander, 48 Conn. 492, 504 (1881) (observing that since the testator "lived nearly five years after the divorce making no changes in his will, ... the conclusion is well nigh irresistible that he did not intend to deprive his former wife of the provision he had made for her."); Trueblood v. Roberts, 732 N.W.2d 368, 376 (Neb. Ct. App. 2007) (Sievers, J., concurring) ("[T]he key collateral fact here is that [the decedent] did not in the 11 years after the divorce ... change the beneficiary on his policy."); Bowdlear v. Bowdlear, 112 Mass. 184, 186 (1873) (making a like observation); Charlton v. Miller, 27 Ohio St. 298, 305 (Ohio 1875) (same); In re Arnold's Estate, 110 P.2d 204, 205 (Nev. 1941) (same); Murphy v. Markis, 130 A. 840, 842 (N.J. Ch. 1925), aff'd 132 A. 923 (N.J. 1926) (same); In re Silberstein's Will, 108 N.Y.S.2d 88, 92 (Sur. Ct. 1951) (same). Cf. In re Estate of McQuay, 335 N.E.2d 746, 750 (Ohio Ct. App. 1975) (testator "died seventeen days after the divorce was granted, during which time he was ill and could hardly be expected to have changed his will. Perhaps this is a portion of the circumstances which need to be noted."). One court contrasted the two situations: "[I]t cannot be said that this testator, who survived for many months after the divorce, lacked the opportunity to revoke the will in a manner provided by [law] ... (cf. Matter of Ga Nun's Estate ... where the testator died one month after the divorce)." In re Estate of Torr, 186 N.Y.S.2d 205, 208 (Sur. Ct. 1959) (citing In re Ga Nun's Estate, 104 N.Y.S.2d 344 (Sur. Ct. 1951)). In other cases, however, proximity to death has not entered into judicial discussions. See, e.g., Gartin v. Gartin, 21 N.E.2d 289, 290 (Ill. 1939) (testator died within thirty days of divorce); In re Cabaniss' Estate, 129 P.2d 1003, 1004 (Okla. 1942) (testator died two months after divorce); Prudential Ins. Co. v. Weathersfield, 621 P.2d 83, 84 (Or. Ct. App. 1980) (insured died six days after divorce). Whereas two courts have affirmatively rejected a rule of implied revocation premised on proximity to death, Bowers, 637 S.W.2d at 457-58 (overruling appellate court which had sought to install such a rule covering the case of divorce); In re Group Life Ins. Proceeds of Mallory, 872 S.W.2d 800, 803 (Tex. App. 1994) (rejecting argument made by appellants), a third court has affirmatively rejected the argument (made by appellant) that a rule of implied revocation should not apply when a testator survived annulment (held equivalent to divorce) for a long time without modifying his will. Johnston v. Laird, 52 P.2d 1219, 1223 (Wyo. 1935).

(145.) Aetna Life Ins. Co. v. Wadsworth, 689 P.2d 46, 52 (Wash. 1984).

(146.) Id. One foreign jurisdiction where a rule of this sort exists today is the Republic of South Africa. By statute in that nation, bequests to a spouse are impliedly revoked upon divorce, but the presumption applies only if the testator dies within three months of the divorce. Wills Act 7 of 1953 s. 2B (amended 1992). "In its report the [South African] Law Commission said that if a will was not amended within three months, it was arguable that the testator intended to benefit the former spouse after all." M.C. Schoeman-Malan, Revocation of Wills by Changed Circumstances, in EXPLORING THE LAW OF SUCCESSION: STUDIES NATIONAL, HISTORICAL, AND COMPARATIVE 141, 153 (Kenneth G.C. Reid et al. eds., 2007).

(147.) WASH. REV. CODE ANN. [section] 11.07.010 (West Supp. 2003). Lawmakers enacted this statute in response to the decision in Wadsworth. On the legislative history, see Mearns v. Scharbach, 5 P.3d 29, 33 (Wash. Ct. App. 2000). But cf. Henley v. Henley, 974 P.2d 362, 364 n.5 (Wash. Ct. App. 1999) (suggesting that Wadsworth may retain some vitality).

(148.) For a mathematical proof of the insignificance of short-term error costs, see the Appendix.

(149.) Four categories of events merit investigation. (1) Changes of relationship: This category is the focus of the current laws of implied revocation. Lawmakers have lighted on marriage and divorce as triggering events, but do other events also change intent along this continuum? What about formal engagement? See supra note 122 and accompanying text. By the same token, what about broken engagement, if the testator executed a premarital will? By comparison, an implicit condition of marriage attaches to a prenuptial gift. RESTATEMENT (SECOND) OF PROP.: DONATIVE TRANSFERS [section] 31.2 cmt. c & illus. 5-9 (1983). What about permanent separation? Law reformers in New York considered applying the doctrine of implied revocation to instances of separation but decided that the case was insufficiently compelling. Samuel Hoffman, Revocation of Wills and Related Subjects, 32 BROOK. L. REV. 1, 32 (1965). In the Canadian province of British Columbia, however, a judicial decree of separation suffices to trigger an implied revocation. Wills Act, R.S.B.C. [section] 16(2) (1996). In a number of states, an abandoning or adulterous spouse forfeits inheritance rights, and in two instances the operative statutes could be read to cover bequests under wills, see IND. CODE ANN. [section][section] 29-1-2-14, 29-1-2-15 (West Supp. 2008); KY. REV. STAT. ANN. [section] 392.090 (LexisNexis 1999), but the case law has construed the statutes more narrowly. Baldwin v. Cook, 23 S.W.2d 601, 601-05 (Ky. 1930); see also Oliver v. Estate of Oliver, 554 N.E.2d 8, 9-11 (Ind. Ct. App. 1990) (issue not raised directly and prior to statutory amendment). By the same token, what about reconciliation following a separation? And should other changes of relationship also operate as triggering events? What about a bequest to a child whom the parent subsequently gives up for adoption (structurally analogous to divorce)? What about a bequest to an employee whose employment subsequently terminates?

(2) Wrongdoing: A beneficiary may harm the testator sufficiently to damage their relationship without formally changing it (and without causing the testator's death, as discussed supra Part I.C.2). Should such harm trigger an implied revocation? In Louisiana, a beneficiary who is judicially determined to have attempted to murder the testator loses his or her bequest under the will, absent a finding of reconciliation or forgiveness. LA. CIV. CODE ANN. arts. 941, 943, 945 (2000). In Oregon, a beneficiary convicted of physically or financially abusing the testator loses his or her bequest under the will if the testator dies within five years of the conviction, but this forfeiture operates as a mandatory, not as a default, rule. OR. REV. STAT. ANN. [section][section] 112.455(1), 112.465(1) (West 2005). See also CAL. PROB. CODE [section] 259 (Deering 2004) (abuser cannot recover via the will any damages stemming from abuse awarded to the testator's estate); McDonald v. Carey, No. A113265, 2008 WL 4383876, at *25 (Cal. Ct. App. Sept. 29, 2008) (construing this statute). Under early English law, implied revocation applied to any beneficiary who "become[s] enimie to the testator." SWINBURNE, supra note 141, at 286. And ought not these principles operate reciprocally--that is, does not a serious harm committed by the testator upon the beneficiary likewise signal a breakdown in their relationship that probably entails a wish to disinherit that beneficiary? See supra text at notes 61-62.

(3) Changes of status: Should a beneficiary's conviction of a crime of moral turpitude trigger an implied revocation? Should a beneficiary's insolvency or bankruptcy petition trigger an implied revocation? In cases of insolvency where a bankruptcy petition and discharge is likely to follow, inherited assets only benefit creditors of the beneficiary; even after bankruptcy, assets inherited within six months of the petition for relief flow back into the bankruptcy estate to satisfy creditors. 11 U.S.C. [section] 541(a)(5) (2006). For discussions raising the possibility of implied revocation in this connection, see Adam J. Hirsch, The Problem of the Insolvent Heir, 74 CORNELL L. REV. 587, 632-36 (1989); Adam J. Hirsch, Inheritance and Bankruptcy: The Meaning of the "Fresh Start", 45 HASTINGS L.J. 175, 237-38 (1994).

(4) Changes of property: Courts have rejected claims of implied revocation on the ground of substantial changes in the value of a testator's property, e.g., Aten v. Tobias, 220 P. 196, 202 (Kan. 1923), although in scenarios where the will evidences an intent to equalize bequests that grow unequal due to fluctuating values, probable intent might be reconstructed with some assurance. See id. (dictum).

(150.) This formulation constitutes a dual standard: both its substantive and temporal coverage are fuzzy. Alternatively, the triggering events could be fuzzy within a fixed space of time; or the triggering events could be fixed, within a fuzzy space of time. The temporal limit on the proposed power distinguishes this rule from a broader power to reform wills on the basis of probable intent. See supra notes 103-04 and accompanying text.

(151.) Testators can opt out of default rules of implied revocation without reference to specific facts. Evans v. Palmour, 553 S.E.2d 585, 587 (Ga. 2001); see also Hirsch, supra note 37, at 1065-69 (on the theory of discretionary default rules).

(152.) "Such proof would be tantamount to allowing parol declaration of intent...." In re Torregano's Estate, 352 P.2d 505, 513 (Cal. 1960). "Declarations made by the testator after the divorce ... cannot ... modify the effect of a state of facts which was at the time sufficient. To allow that to be done would indeed make unstable, if not utterly destroy, the rule itself." In re Martin's Estate, 190 N.W. 872, 874 (Neb. 1922). "One may not dispose of his property by mere intentions ...." In re McGraw's Estate, 207 N.W. 10, 11 (Mich. 1926). In jurisdictions where extrinsic evidence is admissible, however, courts admit declarations into evidence. Fillers v. Cannon, No. 105, 1986 WL 3889, at *1-2 (Tenn. Ct. App. Apr. 1, 1986). Several opinions defend this rule of evidence by analogy to revocation by act, where parol evidence of intent is once again admissible. In re Estate of Mercure, 216 N.W.2d 914, 918-19 (Mich. 1974); In re McGraw's Estate, 207 N.W. at 13 (Sharp, J., dissenting). Here, the misidentification of the instant doctrine as a form of revocation, see supra text at note 110, causes confusion. Declarations pertaining to revocation by act are credible because the additional symbolic action required demonstrates intent to achieve legal consequences. See also supra note 130 and accompanying text. No symbolic act accompanies an implied revocation; all we have is a naked declaration, making its intended significance and finality less clear.

(153.) Even regarding divorce, courts acknowledge that instances in which testators wish to maintain estate plans benefitting a former spouse "are not infrequent," In re McGraw's Estate, 207 N.W. at 12 (Sharpe, J., dissenting), are "a real possibility," Whirlpool Corp. v. Ritter, 929 F.2d 1318, 1323 (8th Cir. 1991), are "not beyond the realm of comprehension," Hinders v. Hinders, 828 So.2d 1235, 1244-45 (Miss. 2002), and a conclusive presumption to the contrary "could lead, in some instances, to an incongruous result," In re Estate of Forrest, 706 N.E.2d 1043, 1047 (Ill. App. Ct. 1999). For a sociological study of post-dissolution relationships, see ANNE-MARIE AMBERT, EX-SPOUSES AND NEW SPOUSES: A STUDY OF RELATIONSHIPS 23-59 (1989).

(154.) In re Estate of Blanchard, 218 N.W.2d 37, 45 (Mich. 1974) (Kavanaugh, C.J., dissenting) (also emphasizing the need for certainty); In re Estate of Knospe, 626 N.Y.S.2d 701, 704 (Sur. Ct. 1995); Mearns v. Scharbach, 5 P.3d 29, 34 (Wash. Ct. App. 2000); see also Luff v. Luff, 359 F.2d 235, 255 (D.C. Cir. 1965) (deeming extrinsic evidence too equivocal).

(155.) Were extrinsic evidence admissible within the framework of a temporary presumption, that presumption could flip after its time expires, so that what was initially a rebuttable presumption of changed intent becomes a rebuttable presumption of unaltered intent. A presumption that flipped from rebuttable to conclusive (or vise versa) is also possible but would want structural consistency.

(156.) For an early example of a will executed in contemplation of imminent marriage, see Charlton v. Miller, 27 Ohio St. 298, 299 (Ohio 1875). For examples of wills executed prior to imminent divorces, see In re Buchman's Estate, 281 P.2d 608, 618 (Cal. Ct. App. 1955) (bequeathing part of the estate to spouse while divorce negotiations were in progress); In re Estate of Thompson, 475 N.E.2d 1135, 1136 (Ill. App. Ct. 1985) (disinheriting spouse); Mercure, 216 N.W.2d at 916 (bequeathing entire estate to soon-to-be former spouse); Cooke v. Cooke, 154 N.Y.S.2d 757, 761 (App. Div. 1956) (bequeathing part of the estate to soon-to-be former spouse, and part to soon-to-be new spouse). In Mercure, the court deemed this evidence significant. Mercure, 216 N.W.2d at 916. In Buchman and Cooke, implied revocation was not at issue.

(157.) This distinction, which courts sometimes consider to establish intent in connection with holographic wills and conditional wills, has not figured prominently in implied revocation cases, although it is noted occasionally. Luff, 359 F.2d at 239 (former spouse argued that testator "kept [will] in a bureau drawer in his apartment where it was found after his death, and that this indicated an intention not to revoke it."); Hinders, 828 So.2d at 1243 (former spouse "emphasizes that [testator] ... kept this will in his possession, readily at hand in his desk drawer."); Rankin v. McDearmon, 270 S.W.2d 660, 662 (Tenn. Ct. App. 1953) (predissolution will was "found among some old and worthless papers of the deceased").

(158.) Several courts have suggested that in this event, an inference of intent to revoke a predissolution will may be unjustified. Whirlpool, 929 F.2d at 1323; In re McGraw's Estate, 199 N.W. 686, 688 (Mich. 1924), appeal from new trial 207 N.W. 10 (Mich. 1926); see also Cooke, 154 N.Y.S.2d at 761 (where testator affirmatively bequeathed to spouse in contemplation of divorce, making their daughter contingent beneficiary if spouse predeceased testator).

(159.) Estate of Reeves v. Reeves, 284 Cal. Rptr. 650, 654 (Ct. App. 1991) (evidence inadmissible to override implied revocation); In re Estate of Pekol, 499 N.E.2d 88, 89 (Ill. App. Ct. 1986) (same); Rasco v. Estate of Rasco, 501 So.2d 421, 424 (Miss. 1987) (evidence admitted to override implied revocation); In re Estate of Perigen, 653 S.W.2d 717, 718-19 (Tenn. 1983) (same); In re Estate of Hoevet, 436 P.2d 540, 541 (Wyo. 1968) (same); see also Offerman v. Rosile, 77 P.3d 504, 506 (Kan. Ct. App. 2003) ("Neither party remarried, and [they] apparently continued to have a close social relationship," the admissibility of this evidence not at issue); Mercure, 216 N.W. at 916 (testator lived next door to, ate his meals with, and transferred his telephone service to the home of former spouse and her new husband, sufficing to override implied revocation); Romero v. Melendez, 498 P.2d 305, 306 (N.M. 1972) (couple contemplated remarriage shortly after divorce, this evidence insufficient to overcome implied revocation); Columbus Bar Ass'n v. Sladoje, 776 N.E.2d 1057, 1058 (Ohio 2002) (collateral suit involving estate of testator where divorce resulted in implied revocation even though couple continued to live together).

(160.) UNIF. PROBATE CODE [section] 2-302(a)(2) (amended 2008), 8 pt. 1 U.L.A. 135 (1998) (giving a share to new children only if the testator had children before the will was executed and made bequests to them, and then only a co-equal fraction of those bequests). The accompanying comment refers to pertinent empirical studies. Id. [section] 2-302 cmt. The original Uniform Probate Code operated differently. UNIF. PROBATE CODE [section] 2-302(a), 8 pt. 1 U.L.A. 322 (1998) (pre-1990 Article 2). Statutes based on the revised Code are presently in effect in eleven jurisdictions. RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 9.6 reporter's note 2 (1999). The Commissioners took a step backward in 2008, however, by removing posthumously born children from the purview of this section of the Code. See supra note 121.

(161.) UNIF. PROBATE CODE [section] 2-301(a). Hence, if a premarital will bequeaths the entire estate to children from a prior marriage, no implied revocation will occur at all, although the new spouse can still claim a minimum elective share (or share of the community estate). The original Uniform Probate Code did not include this qualification. UNIF. PROBATE CODE [section] 2-301(a), 8 pt. 1 U.L.A. 317 (1998) (pre-1990 Article 2). Statutes based on the revised Code are presently in effect in twelve jurisdictions. RESTATEMENT (THIRD) OF PROP. [section] 9.5 reporter's note 1.a.

(162.) Lawrence W. Waggoner, Spousal Rights in Our Multiple-Marriage Society: The Revised Uniform Probate Code, 26 REAL PROP. PROB. & TR. J. 683, 750 (1992); see also RESTATEMENT (THIRD) OF PROP. [section] 9.5 cmt. g, illus. 4 (chaired by the same Reporter, and reproducing this reasoning); UNIF. PROBATE CODE [section] 2-301 cmt. (ignoring the issue). The Reporter adds that the testator "without the advice of competent legal counsel ... could hardly be expected to appreciate the need for a new will that would merely repeat the provisions in her old will for her children." Waggoner, supra, at 750. Here, the Reporter acknowledges the risk of error that in truth is implicit in any time-independent doctrine of implied revocation, although the Code does not respond by abandoning those doctrines. On the contrary, the Code elsewhere expands them. See infra notes 167-68 and accompanying text.

(163.) See Waggoner, supra note 162, at 750.

(164.) Approximately 43% of American marriages now end in divorce. The median duration of first marriages dissolved by divorce is currently 8 years, while the median time between a first divorce and remarriage, among those who remarry, is 3.6 years for women and 3 years for men; women's average age at remarriage is 33. Larry Bumpass et al., Changing Patterns of Remarriage, 52 J. MARRIAGE & FAM. 747, 750, 753 (1990); Lawrence Ganong et al., Divorce as Prelude to Stepfamily Living and the Consequences of Redivorce, in HANDBOOK OF DIVORCE AND RELATIONSHIP DISSOLUTION 409, 410 (Mark A. Fine & John H. Harvey eds., 2006); Robert Schoen & Robin M. Weinick, The Slowing Metabolism of Marriage: Figures from 1988 U.S. Marital Status Life Tables, 30 DEMOGRAPHY 737, 742 (1993). One study remarks that "[w]e tend to think of marital disruption as an event well into the life cycle. However, more than half of recent marital disruptions occurred before age 30, about a third before age 25." Bumpass et al., supra, at 749. Apparently, the Commissioners shared this misconception. That is the risk of relying on intuition instead of evidence. See also Arthur J. Norton, The Influence of Divorce on Traditional Life-Cycle Measures, 42 J. MARRIAGE & FAM. 63, 65-68 (1980) (noting a dramatic drop in average age at divorce in the United States over the 20th century).

(165.) This too is nowadays common. Howard Wineberg, Childbearing After Remarriage, 52 J. MARRIAGE & FAM. 31, 31, 37 (1990). In this scenario, under the Code, if the premarital will bequeathed the entire estate to children from the prior marriage, the doctrine of implied revocation would protect only children conceived with the new spouse, not the new spouse himself or herself. UNIF. PROBATE CODE [section][section] 2-301(a), 2-302(a)(2).

(166.) SUSSMAN ET AL., supra note 3, at 91-95 (Ohio data); Fellows et al., Empirical Study, supra note 116, at 727-29, 732 (Illinois data); Fellows et al., Public Attitudes, supra note 116, at 364-68 (data from five states). For anecdotal illustrations, see Gray v. Gray, 947 So.2d 1045, 1046 (Ala. 2006); In re Estate of Perigen, 653 S.W.2d 717, 718 (Tenn. 1983). These data should be known to the Commissioners, since they cite two of the studies elsewhere in the Code. UNIF. PROBATE CODE [section][section] 2102 cmt., 2-302 cmt.

(167.) UNIF. PROBATE CODE [section] 2-804(b)(1). The original Uniform Probate Code confined the doctrine of implied revocation to the former spouse. UNIF. PROBATE CODE [section] 2-508, 8 pt. 1 U.L.A. 376 (1998) (pre-1990 Article 2). Statutes based on the revised Code are presently in effect in some ten jurisdictions. RESTATEMENT (THIRD) OF PROP. [section] 4.1 statutory note 2.a. See also infra note 173.

(168.) UNIF. PROBATE CODE [section] 2-804 cmt.

(169.) See id. [section] 2-804(b).

(170.) Id. [section] 2-804 cmt.

(171.) Waggoner, supra note 162, at 689-90, 694-95 (footnote omitted).

(172.) See the delightful Russian folk tale recounted in J.M. Balkin, Too Good to Be True: The Positive Economic Theory of Law, 87 COLUM. L. REV. 1447, 1461 (1987) (book review).

(173.) In New Hampshire, the statute distinguishes these two cases: Only alternative bequests to stepchildren (or their descendants) are revoked by implication, along with primary bequests to former spouses, under predissolution wills. Otherwise, bequests to stepchildren or in-laws are unaffected by divorce. N.H. REV. STAT. ANN. [section] 551:13 (Supp. 2002); see In re Estate of Sharek, 930 A.2d 388, 390 (N.H. 2007) (applying statute to validate alternative bequest to former brother-in-law).

(174.) See Bloom v. Selfon, 555 A.2d 75, 77 (Pa. 1989) ("Many would view this [bequest to a stepuncle] as an unusual testamentary scheme...."). Since the divorce rate for remarriages is even higher than for first marriages, Teresa Castro et al., Recent Trends in Marital Disruption, 26 DEMOGRAPHY 37, 39, 44-48 (1989), the problem of bequests to former stepchildren is bound to arise with some frequency.

(175.) For review essays remarking the burgeoning literature, see Marilyn Coleman & Lawrence H. Ganong, Remarriage and Stepfamily Research in the 1980s: Increased Interest in an Old Family Form, 52 J. MARRIAGE & FAM. 925, 925 (1990); Marilyn Coleman et al., Reinvestigating Remarriage: Another Decade of Progress, 62 J. MARRIAGE & FAM. 1288, 1288 (2000).

(176.) Maria Schmeeckle et al., What Makes Someone Family? Adult Children's Perceptions of Current and Former Stepparents, 68 J. MARRIAGE & FAM. 595, 606 (2006).

(177.) Constance R. Ahrons & Madonna E. Bowman, Changes in Family Relationships Following Divorce of Adult Child: Grandmother's Perceptions, 5 J. DIVORCE 49, 53 (1981); Anne-Marie Ambert, Relationships with Former In-laws after Divorce: A Research Note, 50 J. MARRIAGE & FAM. 679, 682-85 (1988); Donald Anspach, Kinship and Divorce, 38 J. MARRIAGE & FAM. 323, 325-27 (1976); Jerry W. Spicer & Gary D. Hampe, Kinship Interaction After Divorce, 37 J. MARRIAGE & FAM. 113, 116-18 (1975) (finding parents-in-law interact more with former daughters- than sons-in-law); cf. Janet Finch & Jennifer Mason, Divorce, Remarriage and Family Obligations, 38 SOC. REV. 219, 22528 (1990) (finding normative approval of mutually supportive relationships with former in-laws).

(178.) Marilyn Coleman & Lawrence H. Ganong, Attitudes Toward Inheritance Following Divorce and Remarriage, 19 J. FAM. & ECON. ISSUES 289, 301 (1998); Annelle Z. Weymuth, Attitudes About Intergenerational Family Obligations Related to Providing Inheritance 15 (May 1997) (unpublished Ph.D. dissertation, University of Missouri-Columbia) (on file with author).

(179.) Coleman & Ganong, supra note 178, at 303-04; Weymuth, supra note 178, at 15-17.

(180.) Candan Duran-Aydintug, Relationships with Former In-Laws: Normative Guidelines and Actual Behavior, 19 J. DIVORCE & REMARRIAGE 71, 73, 78-80 (1993); see also ROBERT S. WEISS, MARITAL SEPARATION 144-47 (1975) (reporting anecdotal evidence); Finch & Mason, supra note 177, at 228-44 (also identifying the amicability of the divorce process as a relevant factor).

(181.) The one published discussion reports the results of a survey limited to affluent testators. It found that 40-41% of respondents believe natural children and stepchildren should be treated equally in the estate plan. Whether responses were limited to subjects who actually had stepchildren is unclear. Kim Porter, The Will of the Wealthy, TR. & EST., Aug. 1999, at 56, 56. See also FINCH ET AL., supra note 77, at 101-03 (British data, finding variability in the treatment of stepchildren).

(182.) Experts aside, lay persons appreciate the potential for such bonding sufficiently for it to be depicted on occasion in popular culture (Cinderella notwithstanding). In the delightful feature film THE RELUCTANT DEBUTANTE (MGM 1958), an upper-class British stepmother (played by the late Kay Kendall) is portrayed as intensely ambitious for her half-American stepdaughter and reacts with raptures of joy when her fiance, at first mistaken for a bounder, is revealed as heir to an Italian duke.

(183.) For recent empirical studies, see Sandra L. Hofferth & Kermyt G. Anderson, Are All Dads Equal? Biology Versus Marriage as a Basis for Paternal Investment, 65 J. MARRIAGE & FAM. 213 passim (2003) (also finding that stepparents' bonds with stepchildren intensify over time); Maria Schmeeckle, Gender Dynamics in Stepfamilies: Adult Stepchildren's Views, 69 J. MARRIAGE & FAM. 174 passim (2007); Maria Schmeeckle, Rethinking the Ties that Bind: Adult Children's Perception of Step, Ex-Step, and Biological Parents 88-99 (Dec. 2001) (unpublished Ph.D. dissertation, University of Southern California) (on file with author) [hereinafter Schmeeckle, Rethinking]; Susan D. Stewart, How the Birth of a Child Affects Involvement with Stepchildren, 67 J. MARRIAGE & FAM. 461 passim (2005) (finding that birth of a child affects stepfamilies and intact families similarly). For reviews of the growing literature, see LAWRENCE H. GANONG & MARILYN COLEMAN, STEPFAMILY RELATIONSHIPS 109-44 (2004); Ganong et al., supra note 164, at 417-21.

(184.) For a recent study, see William Marsiglio, When Stepfathers Claim Stepchildren: A Conceptual Analysis, 66 J. MARRIAGE & FAM. 22 passim (2004). See also, e.g., LUCILE DUBERMAN, THE RECONSTRUCTED FAMILY 52-53, 63 (1975) (anecdotal evidence); ELIZABETH EINSTEIN, THE STEPFAMILY 51, 189-90 (1982) (same); for additional sources, see supra note 183.

(185.) One empirical study finds that fewer former than current adult stepchildren view the former or current stepparent as a "parent," but without breaking down the data vis-a-vis the attributes of the former relationship. Schmeeckle et al., supra note 176, at 603-08. Another empirical study finds an "almost unanimous" preference among stepparents for continued contact with stepchildren in the event of divorce. Mary A. Mason et al., Stepparents: De Facto Parents or Legal Strangers?, 23 J. FAM. ISSUES 507, 517 (2002); see also Rosalind Edwards et al., Biological Parents and Social Families: Legal Discourses and Everyday Understandings of the Position of Step-Parents, 13 INT'L J.L. POL'Y & FAM. 78, 97 (1999) (British study finding the same preference). But see Ganong et al., supra note 164, at 428 ("We know of no research on the outcomes of redivorce.... Anecdotally, we know that former stepparents and stepchildren tend to sever relationships with each other after divorce."); cf. Schmeeckle, Rethinking, supra note 183, at 149-54 (presenting anecdotal evidence based on hypothetical questions put to current stepchildren, which elicited varied responses).

(186.) MARGARET M. MAHONEY, STEPFAMILIES AND THE LAW 129-47 (1994); Jason D. Hans, Stepparenting After Divorce: Stepparents' Legal Position Regarding Custody, Access, and Support, 51 FAM. REL. 301, 302-03, 307 (2002); Marie W. Kargman, Stepchild Support Obligations of Stepparents, 32 FAM. REL. 231, 236-37 (1983); Laura W. Morgan, The Rights, Duties, and Responsibilities of Stepparents to Their Stepchildren: Custody and Visitation, 8 DIVORCE LITIG. 185, 186-88, 191-93 (1996); Susan M. Silverman, Note, Stepparent Visitation Rights: Toward the Best Interests of the Child, 30 J. FAM. L. 943 passim (1991-92).

(187.) Clymer v. Mayo, 473 N.E.2d 1084, 1095 (Mass. 1985); Bowling v. Deaton, 507 N.E.2d 1152, 1154 (Ohio Ct. App. 1986); In re Estate of Kerr, 520 N.W.2d 512, 514 (Minn. Ct. App. 1994); In re Estate of Coffed, 399 N.Y.S.2d 548, 551 (App. Div. 1977), aff'd 387 N.E.2d 1209 (N.Y. 1979); Bloom v. Selfon, 555 A.2d 75, 77 (Pa. 1989); In re Estate of Graef, 368 N.W.2d 633, 635-36 (Wis. 1985); see also In re Walker, 849 S.W.2d 766, 769 (Tenn. 1993) (dictum). The Third Restatement nonetheless asserts that courts construing an existing implied revocation statute limited to the former spouse "should effectuate the purpose of the statute by extending its terms to revoke the devise to [a testator's former stepchildren]." RESTATEMENT (THIRD) OF PROP.: WILLS AND OTHER DONATIVE TRANSFERS [section] 4.1 illus. 12 & reporter's note 10 (1999) (chaired by the Reporter of the Uniform Probate Code). However we assess the Code's expansive approach as a matter of policy, we may harbor misgivings when a Restatement, as a matter of jurisprudence, gives its imprimatur to acts of judicial legislation.

(188.) "There is no reason to assume that in all or even most cases a testator estranged from his spouse will also be estranged from ... relatives of that spouse." Porter v. Porter, 286 N.W.2d 649, 655 (Iowa 1979). "This is a highly individualized matter, ... and it cannot be said that the likelihood of estranged feelings is sufficiently high as to warrant nullification.... Such a [bequest] may evidence an unusually close bond [to the in-law]." Bloom, 555 A.2d at 77. See also Clymer, 473 N.E.2d at 1095; Kerr, 520 N.E.2d at 514 (similar statements). But see Hermon v. Urteago, 46 Cal. Rptr. 2d 577, 581 (Ct. App. 1995) (endorsing the Code and its rationale in a will construction case). Cf. In re Estate of Nash, 220 S.W.3d 914, 916-18 (Tex. 2007) (possibly reflecting judicial bias against a stepchild).

(189.) Cf. Estate of Jones, 18 Cal. Rptr. 3d 637, 638, 643-45 (Ct. App. 2004) (resolving whether a will's description of beneficiaries as "my stepdaughters" imposed a condition that they maintain that status by admitting extrinsic evidence to show whether "the testator had a relationship with a child of a divorced spouse that continued following the divorce," and also deeming relevant the fact that the appellant stepdaughter "was already an adult when her mother married [the testator], and so there is no issue of the bond associated with raising a child.").

(190.) UNIF. PROBATE CODE [section] 2-804(b) (amended 2008), 8 pt. 1 U.L.A. 61 (Supp. 2008). Court orders can override a presumption of revocation, but only by their "express terms." Id. If the testator adopts a stepchild, a bequest to him or her is not impliedly revoked by divorce, id. [section] 2-804(a)(5), but this option may not exist if the nonresidential parent remains alive. MAHONEY, supra note 186, at 16177. For an empirical study, see Kathleen A. Lamb, "I Want to Be Just Like Their Real Dad": Factors Associated with Stepfather Adoption, 28 J. FAM. ISSUES 1162 passim (2007). See also infra note 195 and accompanying text.

(191.) See supra notes 154-59 and accompanying text.

(192.) UNIF. PROBATE CODE [section] 2-804(b). See generally Susan N. Gary, Applying Revocation-onDivorce Statutes To Will Substitutes, 18 QUINNIPIAC PROB. L.J. 83 (2004).

(193.) UNIF. PROBATE CODE [section][section] 2-301(a), 2-302(a). Cf. CAL. PROB. CODE [section][section] 5000, 5600, 6122, 21601, 21610, 21620 (Deering 2004) (applying all the rules of implied revocation to both wills and will substitutes, although not with perfect consistency). See also In re Estate of Jackson, 194 P.3d 1269, 1273-75 (Okla. 2008) (construing non-Uniform implied revocation statute to apply only to wills).

(194.) UNIF. PROBATE CODE [section] 2-301(a) & cmt.

(195.) Id. [section] 2-302(a). This scenario has produced a steady stream of litigation. For cases analogizing the problem of premarital bequests for then-friends with pre-adoption bequests for after-adopted children, see Estate of Turkington, 195 Cal. Rptr. 178, 180 (Ct. App. 1983); In re Guilmartin's Estate, 282 N.Y.S. 525, 526-30 (Sur. Ct. 1935), aff'd 293 N.Y.S. 665 (App. Div. 1937), aff'd 14 N.E.2d 627 (N.Y. 1938). For additional cases on point, see Davis v. Davis, 365 A.2d 1004 (Md. 1976); Bowdlear v. Bowdlear, 112 Mass. 184 (1873); In re McPeak Estate, 534 N.W.2d 140 (Mich. Ct. App. 1995); In re Stier's Will, 345 N.Y.S.2d 913 (Sur. Ct. 1973); Brown v. Crawford, 699 P.2d 162 (Okla. Civ. App. 1985); DeSolms Estate, 8 Pa. D. & C.2d 322 (Orphans' Ct. 1956); In re Estate of Hamilton, 441 P.2d 768, 770 (Wash. 1968).

(196.) UNIF. PROBATE CODE [section] 2-301(a)(1).

(197.) Id. [section][section] 2-302(b)(1), 2-804(b) (also providing that a contract made within the context of a divorce settlement supersedes an implied revocation).

(198.) Id. [section][section] 2-301(a)(3), 2-302(b)(2).

(199.) Id. [section] 2-609. See supra notes 127-28 and accompanying text.

(200.) Id. [section] 2-609(a) (a gift may satisfy a bequest "in whole or in part").

(201.) Id. [section][section] 2-301(a), 2-302(b).

(202.) Id. [section] 2-609(a).

(203.) See supra note 198 and accompanying text.

(204.) UNIF. PROBATE CODE [section] 2-302 cmt., 8 pt. 1 U.L.A. 322 (1998) (pre-1990 Article 2).

(205.) UNIF. PROBATE CODE [section][section] 2-109(a), 2-609(a).

(206.) See supra notes 126, 130 and accompanying text.

(207.) In its attention to the size of the gift, however, the Code's approach to satisfaction of implied bequests bears a closer resemblance to the common law. See supra note 127.

(208.) The general exclusion of extrinsic evidence in connection with implied revocation also contrasts with the Code's wide admissibility of extrinsic evidence in connection with ademption by extinction. See supra note 98. The Code's treatment of stepchildren under its rules of implied revocation likewise clashes with their treatment under the Code's antilapse rule. See supra note 90.

(209.) But cf. Richard Primus, When Should Original Meanings Matter?, 107 MICH. L. REV. 165 passim (2008) (arguing against a monolithic approach to constitutional interpretation).

(210.) William H. Page, Ademption by Extinction: Its Practical Effects, 1943 WIS. L. REV. 11, 38.

(211.) Lawrence W. Waggoner, The Uniform Probate Code Extends Antilapse-Type Protection to Poorly Drafted Trusts, 94 MICH. L. REV. 2309, 2337 (1996).

(212.) Id.

(213.) Id. at 2337-38.

(214.) For fuller discussions, see Hirsch, supra note 37, at 1070-72; William Meadow & Cass R. Sunstein, Statistics, Not Experts, 51 DUKE L.J. 629 passim (2001). A disaffected Commissioner likewise warns that the drafters of the UCC, who "must be satisfied by 'second best' data such as ... the experience of the Reporters, [and] members of the Drafting Committee, ... will frequently be 'shooting in the dark'." Richard E. Speidel, Revising UCC Article 2: A View from the Trenches, 52 HASTINGS L.J. 607, 608-09 (2001).

(215.) See supra notes 90, 164-66 and accompanying text.

(216.) See supra notes 132, 175-86 and accompanying text.

(217.) See supra notes 117, 162.

(218.) E.g., Robinson v. Ward, 387 S.E.2d 735, 738 (Va. 1990) (quoting Bell v. Timmins, 58 S.E.2d 55, 59 (Va. 1950)). But compare the Commissioners, who premise at least one default rule on legal assumptions that a testator "knows (or should know)." UNIF. PROBATE CODE [section] 2-509 cmt. (amended 2008), 8 pt. 1 U.L.A. 154 (1998). Does the obligation upon legal actors to know the law mean (as the Commissioners' statement implies) that lawmakers can discount the importance of error costs when formulating rules initially?

(219.) As an early English critic of implied revocation observed, "It may safely be asserted, that the present law is not relied upon in practice; for no one neglects to cancel or alter his will, because he knows that his marriage and the birth of a child will have the effect of revoking it." SUGDEN, supra note 71, at 204 app.

(220.) In this respect, an error-minimizing default can play a role similar to that of a penalty (or information-forcing) default. For a discussion of the latter, see Ian Ayres & Robert Gertner, Filling Gaps in Incomplete Contracts: An Economic Theory of Default Rules, 99 YALE L.J. 87 passim (1989).

ADAM J. HIRSCH, William & Catherine VanDercreek Professor of Law, Florida State University. M.A. (1979), J.D. (1982), Ph.D. (1987), Yale University. Thanks to Grayson McCouch and William McGovern, Jr., and to auditors at faculty colloquia at Boston College Law School, the University of Connecticut School of Law, George Washington University Law School, George Mason University School of Law, New England School of Law, the University of Pittsburgh School of Law, and the University of San Diego School of Law, and at the Succession Law in the 21st Century Conference, for helpful comments.
COPYRIGHT 2009 Washington University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:Hirsch, Adam J.
Publication:Washington University Law Review
Date:Jan 1, 2009
Words:28098
Previous Article:Unsophisticated wealth: reconsidering the SEC's "accredited investor" definition under the 1933 Act.
Next Article:Friends as fiduciaries.
Topics:


Related Articles
Redemption Agreement Transferring Stock to Business Associate Was Not "Testamentary Device"; 2d Cir. Reverses TC.
Buy-Sell Agreements.
QSSTs.
Tax Planning for Lifetime and Testamentary Dispositions-Prototype Plans. (Books and References: Taxes).
Who what when and why? Capacity decisions.
Testamentary conditions in restraint of religion in the twenty-first century: an Anglo-Canadian perspective.
Defining and supporting entrepreneurial profit, entrepreneurial incentive, and external obsolescence.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters