Printer Friendly

Epigraphic geography: the tribute quota fragments assigned to 421/0-415/4 B.C.

ABSTRACT

In this article, the author examines fragments of the Athenian tribute quota lists assigned to 421/0-415/4 B.C., the period preceding the elimination of tribute in ca. 413. The epigraphical and historical arguments employed previously in the reconstruction of these lists are on the whole not cogent. Moreover, in some cases, epigraphic anomalies such as differences in lettering and the uncertainty of joins may challenge the association of fragments within lists. It is suggested that many fragments could equally well be dated to the period following the increased assessment of 425, a period that currently constitutes the sole gap in the reconstructed tribute record.

**********

The first decade of the Peloponnesian War (431-421 B.C.), the so-called Archidamian War, took its toll on the Athenian treasury. (1) In the years following the Peace of Nikias, concluded in winter 422/1, however, Athens evidently recovered its financial strength--so Thucydides has Nikias claim in a speech before the assembly during a debate about the proposed Sicilian expedition. (2) Yet, by 413, the rosy picture had darkened. In or around that year, increased expenditures and declining revenues drove the Athenians to dismantle their tributary system of obtaining imperial revenue, in place for ca. 65 years, and to substitute in its stead a five-percent maritime tax, an eikoste, in hopes of extracting more revenue from the arche (Thuc. 7.28.4). Whether this radical change came about swiftly or as a result of some years of debate and planning cannot be known. (3) The decision and the fiscal environment that prompted it, however, immediately elevate the importance of the preceding years in the larger history of tribute.

EVIDENCE FOR TRIBUTE ASSESSMENT AND COLLECTION OTHER THAN QUOTA LISTS

Evidence for tribute assessment and collection during this period is both literary and epigraphical. Among the literary sources, Andokides (3.8-9) states that the Athenians collected over 1,200 talents of tribute annually during the period of the Peace, a figure that Plutarch (Arist. 24.3) increases to 1,300 talents. (4) [Andokides] (4.11) attributes to Alkibiades a twofold increase in tribute in 418/7. Scholars have tended to be skeptical of claims like these, given the exaggeration and blatant inaccuracy that color fourth-century oratory--especially when orators enlist fifth-century Athenian history for their rhetorical agenda--and given the problems that the late source Plutarch poses for specific details of fifth-century history. In any case, such sources, without a check, fail to bring us closer to understanding why the Athenians would choose to abolish tribute in ca. 413; on the contrary, they make the decision more difficult to understand. These problems in general bring epigraphic evidence to center stage.

Besides the tribute quota lists, recording the dedication of 1/60th of the tribute of the cities to Athena, the editors of ATL, Meritt, Wade-Gery, and McGregor, placed great weight both on the assessment of tribute in 422/1 (IG [I.sup.3] 77) (5) and on inferences drawn from the terms of the Peace of Nikias as outlined by Thucydides (see below) to test the validity of the claims made by literary sources. According to the editors' historical reconstruction, (6) the assessment of 422/1 reduced, "wherever feasible," the amounts of tribute imposed on cities to their levels at the outset of the Delian League in 478/7 (or commensurate with those levels in the cases of cities that had not been original members of the League), the so-called tribute of Aristeides. (7) In the years following the assessment of 422/1 the Athenians collected ca. 500 talents of tribute each year, until 414/3, when tribute was abolished. By that year, presumably because of the modest intake combined with unusual expenditures, most notably on the Sicilian expedition, the treasury had dropped to its 422 level. According to this reconstruction, the quota lists and the assessment of 422/1 were mutually corroborating, and both in turn provided a crucial corrective to an inaccurate literary tradition; the income from tribute never reached 1,200 or more talents during this period, and no increase in assessment occurred in 418/7. (8)

The view that the assessment of 422/1 represented a dramatic reduction consequently guided the editors of ATL in their assignment of fragments of quota lists to the years following 421. It is important, however, to understand on what this view depends. It owes perhaps most to an inference drawn from Thucydides, who records as one stipulation of the Peace of Nikias that six cities in the north that had revolted from Athens were to be required to pay "the tribute of Aristeides" (5.18.5). Following West, Meritt and his coeditors argued that, if cities disloyal to Athens were to pay at this level, "loyal friends" could not have been expected to pay a higher amount. (9) Yet, in accordance with the strictures embodied in the 425 assessment, the Athenians would have completed the assessment the previous summer. (10) Thus, a new stage was envisioned, in which the assessment, while "introduced" at its proper time, was then withheld until the Peace so that it could take account of the new peacetime situation. (11)

The final buttress of their interpretation came from a section of the assessment containing the Hellespontine district total. This figure is either fully or only partially preserved: on the line on which the total was inscribed (col. IV.13), the stone breaks off at the beginning of the numerals. If what is preserved on the fragment is the complete figure, then the total was some 96 talents. If, however, numerals originally preceded those preserved, the total would have been 196 talents, if only one numeral was missing, or 296 talents, if two were missing. The editors of ATL preferred the lowest figure, on the grounds that it was more in keeping with the "Aristeidean character of the assessment of 421." (12)

This reconstruction of the historical context may be appealing--among other things it would make the Athenians appear benevolent, now that Kleon was dead--but how secure is it? (13) The inference from the terms of the Peace of Nikias pertaining to the six recalcitrant cities, namely, that all cities would have been extended the same generosity, would not be unreasonable if there were evidence that the Athenians saw it as in their interest to be mild with respect to the allies' obligations. That the Athenians ever had an interest per se in being kind to their allies is a modern assumption, however, unsupported by ancient evidence. (14)

Moreover, it should be remembered that the possibility of a broad extension of the "tribute of Aristeides" depends on manipulating the timing of the assessment, for which there is no evidence whatsoever. Leaving moral judgments aside, it is an equally valid inference that the Athenians in 422/1, needing to replenish their treasury after years of warfare, would have chosen not to offer a large-scale reduction in tribute, especially since their ability to collect (and to enforce collection) would have been enhanced by peacetime conditions. Most important, as Meiggs notes, it was by no means clear in the summer of 422, when the assessment would have occurred, that the Athenians and Spartans would be concluding a peace treaty in the following winter. (15)

Finally, the preference for the smallest figure possible on the fragment containing the Hellespontine total requires that the mason indented the total by two spaces in the column. This is not impossible, but there are no parallels for the practice. A single indentation is, however, precedented. As Meiggs notes, this was the format of the mason who inscribed district totals in the 425 assessment decree, and he regards this as most likely to have been the case in the 422/1 decree, thus making the total 196 talents; extrapolating from that figure, the annual income from tribute would have reached ca. 1,000 talents, thus not far off Andokides' figure of (more than) 1,200 talents. (16) This is certainly as reasonable, if to a slightly lesser degree, as supposing no indentation, which would bring the total to 296 talents (and would mean that the assessment would not constitute a reduction from the 425 assessment). (17)

In any case, so little of the assessment of 422/1 survives--only 30 names of cities are complete or incontestable from what is preserved, with quotas attached to only 23 of these--that it is unwarranted to use this document as strong support for any position, and one needs to exercise extreme caution if using it as a guide in assigning fragments of lists to the following period. Indeed, it is worth asking whether the remains of this assessment decree are even from the 422/1 assessment, since no part of the decree preceding the list of cities and their assessments is extant. (18)

THE TRIBUTE QUOTA LISTS

The range of estimates for the annual income from tribute from ca. 500 to ca. 1,000 talents and the problematic relationship of the literary and epigraphic evidence serve to underscore the importance of the quota lists. How secure is the current arrangement? This question pertains to a larger issue than simply the lists for the period under scrutiny here. Accordingly, the following examination is intended to pinpoint methodological principles and issues that apply to the positioning of fragments of lists in the entire series.

Since the appearance of ATL, scholars have not subjected the systematic arrangement, or "geography," of the fragments as a whole to serious scrutiny, except to question the placement of certain lists or to seek out the year of the notorious "missing list." (19) The explanation for this reluctance lies in the authority of ATL and the daunting nature of the epigraphic remains. Indeed, it is important to appreciate the extraordinary difficulty of the task of making sense of several hundred fragments of quota lists. Meritt himself often rightly expressed ultimate uncertainty over the placement of fragments in lists. Over time--perhaps unavoidably given the sheer number of fragments and the extent of reconstruction--that uncertainty slid into certainty. One can see that process at work between ATL 1 and 2. In the preface of ATL 1, the editors wrote that "this book makes no pretense to being a final publication. In view of the many improvements made in the texts of the tribute lists during the last decade and a half, it would be pessimistic to suppose that further improvement will not be made. We hope that the present summary will serve to consolidate the position so far won and offer a firmer basis for whatever study may be undertaken in the future." (20)

There is no more exemplary statement than this of epigraphists' proper attitude toward epigraphical conclusions--and historical conclusions based on epigraphic evidence. Such a statement is especially apt when applied to evidence such as the quota lists, which are fragmentary, undated, and often extremely difficult to read. In ATL 2, the editors demonstrated the truth of their prediction by confirming the reality of such hands-on epigraphic research, namely, that it is an ongoing process necessitated not only by new evidence but also by new autopsy and the reassessment of arguments for arranging fragments and for the date. In that volume they presented the same fragments again, making a number of changes in readings and datings (though offering no supporting epigraphical commentary). As they wrote, "the restudy of the lists themselves . . . has made advisable first of all a new presentation of the basic evidence." (21)

Certainly the editors felt greater confidence in their new presentation than they did in ATL 1. They did not, however, include any comment to the effect that the evidence presented in ATL 2 could more securely serve as the basis for further study and reappraisal where warranted. Instead, the absence of such a comment tacitly served to make ATL 2 the definitive treatment that contained "facts" about tribute that henceforth could confidently be used in historical reconstructions. Yet the need not merely for reexamination of a particular list but rather for a comprehensive reexamination of all of the lists has not gone unrecognized. (22) To put it bluntly, the certainty with which the history of tribute--and by extension the arche--has been written may be illusory, and that history cannot be written properly until all the fragments have been carefully restudied.

Such a monumental undertaking, if it happens, will take years to accomplish. Seen against that larger scope, the aim here is decidedly modest: to take as a case study the lists attributed to the years of the Peace, and to examine and assess the epigraphical and historical grounds for associating fragments and arranging them as lists during this period. Given that the kinds of considerations and arguments adduced for situating fragments in these years should be regarded as typical of the overall approach to the fragments, it is hoped that this study will make clear the need for caution when using the tribute quota lists generally.

In the following pages I argue that, because of the shifting history of the fragments placed in these years, their association with one another, and questionable epigraphic and historical assumptions that have resulted in the confinement of the fragments to these years, the current placement of lists and groupings of fragments within lists should be considered only tentative and in some cases arbitrary. To return to the historical importance of the lists for the Peace of Nikias in particular, of course, even with accurate knowledge of the total of tribute collected annually in this period, we still would not know the reason(s) for the decision to eliminate tribute. Nevertheless, any attempt at a plausible reconstruction of the historical context must begin with the greatest possible confidence in the epigraphic evidence for tribute collection during the Peace, however lacunose it is; and uncertainty itself constitutes an important kind of knowledge. While it would be incautious at this stage to propose epigraphical and historical reconstructions different from those presented in ATL, I consider here the possibility that many fragments might be repositioned during the period 425/4-422/1, a period in which no lists currently reside.

THE TRIBUTE QUOTA LISTS, 421/0-415/4 (IG [I.sup.3] 285-290)

In IG [I.sup.3], under the current disposition of the quota lists edited by Meritt and McGregor, fragments of lists are assigned to all but one year (419/8) for the period 421/0-415/4. For only one other year, 420/19, does a question mark appear after the date. As a general principle the editors of ATL avoided placing fragments of lists in the period 425-422 because the tribute calculated from the quotas was, with some exceptions, (23) lower than the tribute assessed in 425. Such an assumption, however, should be subject to scrutiny and reevaluation.

List 34: 421/0 (IG [I.sup.3] 285)

List 34 consists of three, nonjoining fragments: frr. 1 (EM 6764), 2 (EM 6758), and 3 (EM 6652). One of these fragments (fr. 1, with its right and top edges preserved) contains roughly the right half of a prescript, including, fortunately, an archon name (Aristion) and dating formula [TEXT NOT REPRODUCIBLE IN ASCII.]. (24) This is the only fragment for the period with a preserved archon name, and accordingly the only fragment that can be dated with certainty, to 421/0, the year in which Aristion held office. In addition to part of the prescript, fr. 1 preserves parts of place-names from what would have been col. II of the stele, and completely intact placenames and quotas for four cities, with part of the name and quota for an additional city; the fragment then breaks off. Two other fragments (frr. 2, 3) have remained with it, fr. 2 situated some 50 lines beneath the end of fr. 1 in the vertical position of what would have been col. II of the stele, and containing part of a Hellespontine panel (with no edges preserved), and fr. 3, with its left edge preserved, and containing names and quotas from an Ionic panel, positioned in col. I.

The association of all three fragments has gone unquestioned; neither fr. 2 nor fr. 3 has been shifted to other years. The fragments may indeed belong to the same list, and therefore all with certainty to 421/0, for the lettering and general appearance of the fragments are similar. Nevertheless, their reverses contain some unappreciated features that add interest to the association (Fig. 1). (25)

[FIGURE 1 OMITTED]

The reverse of fr. 1 is dressed, though not polished like the obverse, and on it an inscription of five lines extends to slightly less than half of the width of the fragment, which itself appears to be roughly half the width of the original stele (Fig. 1, top). (26) It contains a short rubric ("these [cities] paid the army"), followed by three ethnics: Hephaisties, Imbrioi, and Myrinaioi. The right half of the fragment on this side is uninscribed. It seems likely that the entirety of at least this upper part of the stele would have been similarly dressed. It is therefore reasonable to suppose that approximately three-quarters of the width of the stele was uninscribed. Moreover, it is clear that the stele would have been dressed below the extant portion, since at the middle of the bottom edge of the fragment, after a vacat of ca. four lines, is part of a nu and the tops of letter strokes compatible with IA. These letters (dotted in IG [I.sup.3]) are likely to be the end of the beginning of a rubric like that at the top, which would have continued onto the next line and been followed by a list of place-names, though there is no way of knowing how many lines it would have covered.

This opisthographic fragment and the extent of the dressing on the back raise a number of questions that bear on its association with the other two fragments. What is the relationship between the inscriptions on the front and back? Was the back dressed at the same time as the front, or later, when needed? If at the same time, was it in anticipation of immediate or future need for inscribable space? Was the uninscribed part of the reverse of the stele initially dressed because it had been intended for inscription as well? If the back was dressed at the same time as the front, but not inscribed until later, what factors determined how much of the back to dress? Finally, if dressed later, why was more of the surface dressed than was necessary?

It may seem odd to ask these questions, some of which are unanswerable. But the sole reason that questions even arise about the extent of the dressed surface on the reverse of fr. 1, and the presence of dressed but uninscribed areas, is that this fragment is the only one of the three associated fragments that has a smooth surface on its reverse: fr. 2, placed in the center middle, is undressed, as is fr. 3, positioned in the lower left; fr. 3, however, is also drafted along its preserved left edge. (27) Indeed, no one would ever have associated fr. 1 with frr. 2 and 3 on the basis of the reverses.

These anomalies by themselves do not invalidate the association of the fragments, since parallels exist for partly dressed stelai and dressed but partly uninscribed stelai. (28) They do, however, need to be appreciated more fully than they previously have been, since they may cast a shadow of doubt on the association. Moreover, as in the case of all nonjoining fragments, it is not valid to draw firm conclusions about any one fragment on a hypothetically reconstructed stele by using others that have been associated with it. Thus, while it is certainly possible that the three fragments are correctly associated, we cannot state as a fact that the stele was only partially dressed on the reverse: if frr. 2 and 3 do not belong with fr. 1, then the entire reverse of the stele to which fr. 1 belonged may have been dressed.

Whenever the reverse was dressed, if it was done with knowledge of roughly how much space was required, the explanation for so much dressed but uninscribed surface, as in the case of half of fr. 1, is unclear unless the choice was made for aesthetic reasons. But equally, if appearance dictated the decision and frr. 1-3 are correctly associated with each other, why would the whole of the reverse not have been dressed? Moreover, if it was unclear at the time of dressing how much space would be required (i.e., if it had been dressed in expectation of future inscribing), why dress only part of the stele, and what dictated how much to dress? (29)

It would help to ascertain the temporal relationship between the inscriptions on the reverse and the quota list on the obverse. The presence of the preserved top edge of fr. 1 and the absence of a regular prescript preceding the inscription on the reverse, which is introduced simply by the rubric "these (cities) paid the army," warrant the inference that the list in the inscription on the reverse continued one on the front. (30) Dated to the year 421/0, this inscription has special interest, since it would appear that military activity on the islands of Lemnos and Imbros necessitated the rubric, (31) yet we hear nothing from other sources about any conflict in the Hellespontine region at this time, the first year of the Peace of Nikias, and do not hear of any until the Ionian War, following the collapse in Sicily. That no source speaks of any conflict is no objection, since Thucydides, for example, is highly selective in his choice of which military engagements to relate. Moreover, this force might simply have been a regular, established patrolling fleet.

Nevertheless, it is not impossible that the reverse inscription is separated temporally from the obverse. (32) We must leave open both possibilities: that the conflict that led to the use of tribute to pay soldiers in an army or sailors in the fleet occurred at the outset of the Peace of Nikias; or that the reverse preserves part of a list beginning farther down on the obverse, in which case the reverse inscription is not inevitably tied chronologically to that preserved on the obverse. In sum, while the obverse of fr. 1 must be fixed to 421/0, the association of the remaining two fragments must be left unconfirmed, and the temporal relationship between the obverse and reverse inscriptions left open.

List 35: 420/19? (IG [I.sup.3] 286)

List 35 raises different issues from those that pertained to the fragments assigned to the previous list. To make sense of the present location of this list, the history of the positioning of the fragments is crucial (a point that applies generally to all the lists). The shifting of particular fragments from one year to another and the changing association of fragments can mean that the original basis for locating a particular fragment in a given year is no longer operative and therefore it can reveal, often glaringly, the unsatisfactory grounds for placing fragments in a given year, even when the date is accompanied by a question mark. There is no better example of the problem that can arise from the migration of fragments than what is assigned tentatively to 420/19 in IG [I.sup.3]. In ATL 1 three tiny fragments were associated and placed in this year: frr. 1 (EM 12789), 2 (EM 6650), and 3 (now lost). (33) Those assigned as frr. 2 and 3 were subsequently dissociated and appear currently as frr. 3 and 4 of IG [I.sup.3] 289 under the year 416/5. (34) This change left fr. 1 by itself in ATL 2, accompanied by the comment that "this fragment cannot be precisely dated." (35) Thus, this "list" consists of one tiny fragment, which contains portions of the ends of numbers and the letters ON below them. (36)

In ATL 1, the editors argued that the list (then containing the three fragments) could not precede 425 because of the arrangement of letters and numerals on fr. 1. (37) This argument involves (1) the restoration [EI]on[ikOw] from the letters ON that appear at the bottom of the fragment; (2) the assumption that the portions of numbers above it are part of an Island panel; (38) (3) the view that the Island panel did not precede the Ionic panel prior to 425 (which necessitates placing fr. 1 in a year following 425); and (4) the exclusion of the period 425-421, on the basis of a restored quota in line 3 of the fragment, which could not be made to correspond with the quota of any city grouped as part of the Island district panel in the 425 reassessment decree. (39)

Let us take these in order. First, the restoration is justified, for, although [[TEXT NOT REPRODUCIBLE IN ASCII.]] is possible epigraphically, it is vitiated by the closeness of the left edge of the fragment, which would leave insufficient room for the letters to the left of the preserved ON. The assumption that the scraps of numerals above the two letters is from the Island panel, however, is groundless since nothing in the numbers above the evident panel heading indicates to what panel it would have pertained; no beginnings of figures survive on this fragment, only snippets of the ends of several totals. The connection between the 425 reassessment decree and fr. 1 depends on a circular argument: fr. 1 only relates to the decree if it can be shown that the numerals on it belong to an Island panel, but the decree is used to prove that they did and therefore fr. 1 belongs to a list dated after 425.

In the final stage of the argument the period 425-421 was ruled out for fr. 1 because of supposed conflicts between possible restorations of the amounts of quota in the fragment and the totals of tribute that appear in the 425 reassessment decree for cities in the Island district. Simply put, this argument fallaciously equates assessment with collection. It is by no means clear that the cities paid their new assessment in its entirety each year--especially, for example, at times when the allies perceived weakness on the part of Athens. Therefore, nothing prevents this piece from belonging to a quota list dating between 425 and 421. (40)

Thus, EM 12789, a fragment that by itself yields no usable information whatsoever, occupies its own, tentatively dated, year as List 35 because it was initially combined with two other fragments. Upon their removal, it remained as the lone occupant of 420/19?. It is doubtful that it would by itself have been placed in any year and not have been left floating. As mentioned above, even Meritt had reservations about the status of this fragment as List 35, after the removal of the other two fragments. (41) It is not impossible that this fragment could belong to the quota list for the year 420/19, but as it stands there is no justification for thinking that it does, nor that it necessarily should be placed in any year during the Peace of Nikias in preference to another period. If sound arguments can be adduced that demonstrate that fr. 1 contains part of an Island panel, then, for example, the fragment might belong to a list for a year between 425 and 421, or for a year following 413, if tribute collection was at some point resumed. (42)

The assignment of EM 12789 to 420/19? is perhaps the best testimony to the methodological pitfalls that can result from the movement of fragments, arguments from restoration, and historical assumptions. Accordingly, we cannot insist that we possess any part of a list for that year. Under the current ordering, two years should be seen as having no lists: 420/19 and 419/8, a year in which no fragments are currently situated. That no fragments have been assigned to that later year, while EM 12789 is placed in the earlier year, well reveals the arbitrariness of the chronological and dispositional issue in its final state.

List 37: 4 1 8/7 (IG [I.sup.3] 287)

List 37 contains five fragments, three of which--frr. 1 (Agora I 4809), 3 (EM 12798), and 5 (Agora I 7397)--are said to join from behind (Fig. 2). (43) Three fragments preserve parts of a prescript (fr. 1, a substantial middle portion; fr. 2 [EM 6784], several letters from the last two lines from the left portion of a prescript; and fr. 3, part of the bottom line from the right portion of a prescript). (44) Fr. 1 also contains the first four letters of the Hellespontine panel and parts of three rows of numerals below it; frr. 3-5 are positioned in col. II. (45) If indeed fr. 2 is part of the same prescript as fr. 1, we can have greater confidence in the line length than is the case for List 39, examined below. (46) While no archon name is preserved, it would plausibly have been seven or eight letters (so ATL); (47) yet this is not inevitable, since there are no preserved edges and even in the case of stoichedon inscriptions, line lengths can vary. To cite just one example, the right edge of the prescript preserved on fr. 1 of IG [I.sup.3] 285 reveals different line lengths.

[FIGURE 2 OMITTED]

In the process of assigning a date to these associated fragments, the editors of ATL excluded consideration of the years 425/4-423/2 on the grounds that the line length of the prescript prohibited the restoration of names that would have appeared in the heading if the list belonged in one of those years. The conclusion that the list should postdate 423/2, however, did not lead to any certainty about the correct date, for the list has seesawed dizzyingly between 422/1 and 418/7, with Alkaios or Antiphon restored as archon, respectively. In ATL 1, 422/1 was preferred because of preserved figures that suggested a tribute more in accord with the "character" of the 425 assessment; on the basis of its quota in this list, Kyzikos, for example, paid 20 talents in the year to which frr. 1 and 3 belong. (48) Meritt also had argued for 422/1 over 418/7 on epigraphic grounds (involving the restoration of possible names in the prescript). (49) In the time between ATL 1 and 2, however, he reassigned the fragments to 418/7. (50) Yet in ATL 2 they appeared to have returned to 422/1. As Meiggs noted in his review of the volume, "by a slip the archon for 418/7 has been retained in the prescript and one of the main difficulties of the earlier dating has been obscured by the omission of two vacant spaces at the end of line 2." (51) He continued: "On this dating and on the other changes in the lists judgement must be suspended; discussion is reserved for the final volume." (52) In that volume (ATL 3) the editors continued to favor 422/1 on the basis of the high quotas in the list, which they concluded would be out of place in the period 421/0-418/7 on the basis of comparison with other fragments of lists. (53)

The final reassignment came in 1972, when Meritt moved the fragments back to 418/7 on the basis of the new fragment (5), said to join frr. 1 and 3. The new fragment produced part of the Hellespontine panel, and Meritt argued that, with two exceptions, the quotas "indicate[d] a return to the pre-war level of assessment." (54) It should be noted, however, that no quotas are preserved in their entirety, and portions of the ends of figures corresponding to only three of eleven place-names are preserved. The quotas of the two exceptions cited are entirely restored. (55) Most important, if there are quotas that are both commensurate with and low compared to cities' 425 assessments, then it seems little short of arbitrary to fix on 418/7 (weighting the lower quotas) instead of 422/1 (weighting the higher); both would appear to be equally possible, if no other arguments can be offered that are persuasive, even if not cogent.

The variety of arguments adduced in the case of the two years under consideration calls for caution and reserved judgment. Interestingly, Meritt initially raised 412/1 as a possibility, restoring Kallias as the archon, an individual whose demotic matched that preserved on fr. 1, Skambonides (line 2). (56) He immediately rejected it, however, on both epigraphic and historical grounds. Noting that if Kallias were restored as archon, the numeral required for the year of the office of the Hellenotamiai would be [TEXT NOT REPRODUCIBLE IN ASCII.], he argued that this would "exceed by one letter space the maximum amount of stone available even if the word [[TEXT NOT REPRODUCIBLE IN ASCII.]] in line 3 is written without the rough breathing." (57) Yet, as mentioned above, the editors of ATL had to suppose two vacant spaces on the stone with the restoration of the archons Alkaios and Antiphon; the restoration of [TEXT NOT REPRODUCIBLE IN ASCII.] and [TEXT NOT REPRODUCIBLE IN ASCII.] in line 2 would fit the estimated line length exactly.

More serious, in Meritt's view, were objections on historical grounds, namely, that the Athenians abandoned the collection of tribute around 413 (Thuc. 7.28.4). Meritt himself, however, three years earlier had dated five nonjoining, undated fragments, evidently from a reassessment of tribute, to the year 410, arguing that the Athenians had resumed the collection of tribute. (58) From a methodological standpoint, why is it not equally possible that the fragments belong to a list for the year 412 and would thereby reveal the resumption of tribute collection by that year? (59) The use of five nonjoining, undated fragments of an assessment as the mainstay of a significant historical interpretation (that tribute was reimposed in or by that year) and, by contrast, the rejection of 412 as the date of the list to which quota fragments belong on historical grounds (that in ca. 413 tribute was replaced with a maritime tax with no evidence that it was restored in or by 412) create an impression of an arbitrary approach to undated epigraphical evidence.

These fragments, therefore, cannot be said to occupy a secure place in 418/7 as implied by their appearance in IG [I.sup.3] without a question mark. Moreover, there are further epigraphical considerations that may call their association into question. The lettering on the fragments differs in significant respects. Fr. 2, in col. I, and fr. 3, in col. II, have distinctly different nus (see Figs. 3, 4). (60) Likewise the nus in fr. 5 (Fig. 5), positioned on col. II, differ from those on fr. 2, but also, if less pronounced, from those on fr. 3, with which it is said to join. Fr. 1, containing part of a prescript, some numerals, part of a panel heading and the beginnings of some ethnics from col. II, has roughly similar nus to frr. 2 and 3, though they are considerably more varied (Fig. 6). (61) Moreover, the sigmas on fr. 1 differ rather dramatically from those on fr. 3. (62) Since frr. 2 and 3 are positioned in different columns (I and II, respectively), one could suppose two masons at work, one inscribing the prescript and col. I, and another inscribing col. II. (63) In this case, however, we would need to suppose that one mason inscribed the prescript and possibly col. II, and another inscribed col. I, or, alternatively, that three different masons were at work. These epigraphical considerations raise a question about the association, if only about that of fr. 2 with frr. 1, 3, and 5, if those fragments do indeed join.

[FIGURES 3-6 OMITTED]

In sum, with no preserved dating formula, it is possible, but by no means mandated, that Agora I 4809 and the fragments associated with it date to 418/7, or even to another year during the Peace of Nikias. Moreover, it should be left open as to whether the fragments should all be associated.

List 38: 4 17/6 (IG [I.sup.3] 288)

List 38 consists of one fragment (EM 6751) preserving nine full or partial names from an Island panel. (64) It was originally placed in 420/19 or 419/8 (List 35 or 36) in ATL 1, then moved down to List 38 in ATL 2, although the editors also held open the possibility that it belonged to List 37. This shift left no List 36. Even in the list's original location (420/19 or 419/8), the editors of ATL commented that "because of the inconclusive criteria at our disposal the date proposed must be considered tentative." (65) In 1941, Meritt had commented that "from its similarity to List 39 the inscription probably belongs in this assessment period." (66) The downshift from ATL 1 to ATL 2, therefore, seems to have been a result of the reassignment to 416/5 of the two fragments originally associated with the tiny fragment that now comprises List 35 (420/19?). (67) The repositioning, in other words, had nothing to do with a connection between the fragment and the year 417/6.

In addition, calculated tribute totals from the fragments now assigned to 418/7, examined above, which had previously occupied a reconstructed list for 422/1 in ATL 1 and 2, corresponded more closely with those from EM 6751. Thus, EM 6751 in a sense accompanied its relatives to a later date. Both lists, that for 417/6 and 416/5, were believed initially to reflect a new, higher assessment of 418; no assessment decree, however, survives from which to derive support for this new placement. Indeed, the tribute totals for some of the cities in this list would, like some in 418/7, not be out of place in the period 425-421. For example, inferring from this fragment, Kythnos in this year (whenever it was) paid six talents, which was its assessment in 425, having previously in 432 paid three. Keioi also paid six talents; it was assessed ten in 425, and paid only three in 432. The assessments of the other cities listed in this fragment are not extant.

The current placement of EM 6751 in 417/6 on the basis of its compatibility with the fragments placed in 416/5 depends ultimately on the validity, if not cogency, of the arguments for positioning fragments in 416/5. They are the most elaborate of those advanced for dating lists in the period under scrutiny here. Let us now examine the case.

List 39: 416/5 (IG [I.sup.3] 289)

List 39 contains four fragments, including one long lost. Parts of (evidently) personal names (or, e.g., demotics) are preserved on one (fr. 1, EM 13048), positioned as part of a prescript of a quota list (Fig. 7), and on another (fr. 2, EM 13049), the end of a prescript of a quota list, followed by a list of names from the Island panel with some quotas preserved (Fig. 8). Another fragment (fr. 3, EM 6650) preserves only the beginning letters of two names, with quotas completely or partially preserved. The lost fr. 4 preserved names, also with some quotas completely or partially preserved.

[FIGURES 7-8 OMITTED]

In ATL 1, frr. 3 and 4 were thought, on the grounds of their similarity to those in List 34, to be from 421/0, and certainly earlier in the Peace rather than later (based on an assumption that tribute was increased in 418/7). (68) They were originally combined with EM 12789, the tiny fragment that occupies List 35, examined above. In ATL 2 they were moved to 416/5, where they are now part of IG [I.sup.3] 289, because of their resemblance to two new fragments from the Nike bastion (EM 13048 and 13049) discovered in the interim between ATL 1 and 2 and published by Welter (Figs. 7, 8). (69) Arriving at the new date involves a series of complex stages. Most of the supports for the argument that these fragments all belong together, and, especially, that they date to 416/5, however, are far too shaky to carry the weight placed on them.

The argument begins with the association of EM 13048 and 13049, and from there moves on to an estimate of the distance separating the two fragments. It ends with the assignment to List 39 (416/5) of all four associated fragments, through the restoration in fr. 2 (EM 13049) of the name Arimnestos, archon in 416/5. (70) Let us first consider the proposed association. Although Welter did not think that the two fragments came from the same stele, Meritt argued, on the basis of a squeeze and J. H. Oliver's concurrence from autopsy, that they did, and that, while there was no physical join, they could be arranged in such a way as to position fr. 1 at the beginning of a prescript and fr. 2 at its end. The main part of the argument, beginning with the assumption that fr. 1 must be part of a prescript of a quota list and belongs with fr. 2, concerns the restoration of the name Arimnestos, which involves highly intricate reconstructive footwork. It depends above all on a crucial letter in the top line of fr. 1 and necessitates the rejection of one possibility for the letter, an argument about prescript headings, assumptions about the relative positioning of the two fragments and the line length of the prescript, and a restored prescript heading that has no parallel in any quota list.

On the top line of fr. 1, Welter read AR and before it a trace of a lower vertical stroke. Meritt, from a squeeze, seconded by Oliver from autopsy, saw part of a stroke of an iota compatible with Welter's reading, and before that the vertical and top horizontal bars of either an epsilon or a pi. Situating this fragment at the beginning of a prescript, missing only one initial letter, he then considered and rejected the possible restored formula [[TEXT NOT REPRODUCIBLE IN ASCII.].] on the grounds that quota lists from this period do not begin that way, but rather with the name of the secretary of the boule, followed by the archon formula [TEXT NOT REPRODUCIBLE IN ASCII.]. Meritt also objected to the length and wordiness of possible restorations of the full prescript (combining frr. 1 and 2) compared to others of the period (IG [I.sup.3] 285 and 287).

The next stage of his argument is the crux, and the point where possibility moves swiftly to certainty. As Meritt put it, "there is an added epigraphic objection which I hesitate to press to its full and logical conclusion because the surface in line 1 is so badly weathered as to make absolute certainty of the reading questionable." (71) He then noted that, while the top horizontal stroke of pi "elsewhere in the text" measures 0.009 m, the stroke in line 1 "seems to measure only 0.007 m in length." (72) He thus read epsilon instead of pi, which allowed the restoration [TEXT NOT REPRODUCIBLE IN ASCII.]. This restoration necessitated lengthening the distance between the two fragments (from 65 to 72 letters), and led to the conclusion that the archon dating formula could not have occurred in line 1 because of the posited line length, but instead could be placed in the lower part of the prescript. This, he argued, would be possible if the phrase ending the prescript, [TEXT NOT REPRODUCIBLE IN ASCII.], "is given the compendious form" that appears in IG [I.sup.3] 285 and 287.

The argument culminates with the observation that "surely it is more than mere coincidence that where the name should fall, in the line above the letters [TEXT NOT REPRODUCIBLE IN ASCII.], the preserved letters [SIGMA]TO form part of the archon-name [TEXT NOT REPRODUCIBLE IN ASCII.], one of the few that are available for restoration. The appearance of this name dates the inscription in the year 416/5." (73) He then offered a new reading of the prescript, (74) as follows:

5 [TEXT NOT REPRODUCIBLE IN ASCII.]

The association of frr. 1 and 2 is possible--if only because the poor condition of fr. 1 allows it to be associated with practically anything. (75) In short, the association of frr. 1 and 2, as well as the additions of fr. 3 and the lost fr. 4 to the same list as frr. 1 and 2, cannot be proved or disproved. At the same time, however, several points are crucial to appreciate when assessing the strength of the case for the association of frr. 1 and 2. First is the severely weathered condition of fr. 1, which contains the letter stroke at the heart of the argument. What is certain is that, from the juxtaposition of letters below line 1, the inscription clearly contains parts of personal names (or, e.g., demotics). What is uncertain, and is a critical caveat to the entire reconstruction, is that the top edge of the fragment is a portion of the top edge of the stele to which it belonged, as is maintained. (76)

Second, and the most serious objection to the argument, is the conversion from a tentative possibility to fact of the epigraphical point Meritt "hesitate[d] to press" but then immediately pressed, namely, the length of the top horizontal bar of a letter, and the methodological unsoundness of the support that he adduced for it. Even if he correctly read the complete top bar of a letter, the condition of the stone's surface does not permit precise measurement. (77) In any case, it is not legitimate to argue for a particular letter on one fragment by drawing support from a letter on another, nonjoining fragment. In short, the assignment of four fragments to the year 416/5 depends on a highly tenuous reading of a single letter and on a problematic use of nonjoining fragments. The authority conferred on this one letter on the most weathered line of a nearly illegible fragment finds no better testimony than its appearance in IG without an underdot. (78)

Finally, even if epsilon is the correct reading, in the letter space to its right, where Meritt read part of an iota, are traces of what appears to be a sigma. (79) If this is the correct reading, which on several examinations it seemed to be, then both Welter's and Meritt's suggested restorations are invalidated. Readings aside, however, it must be regarded as fundamentally uncertain whether this fragment even preserves part of a prescript of a quota list, for the first legible line, restored as part of a dating formula of a prescript, could preserve part of another personal name (or demotic, etc.) instead; thus, the fragment may have come from an entirely different kind of inscription. (80) The fact that the two fragments were found in proximity to each other is not a compelling argument for their association with one another. As mentioned above, Welter, who published the results of the excavation of the Nike bastion that produced them, thought that they did not belong to the same inscription.

It is instructive to underscore the additional methodological problems with Meritt's argument, noted above, that the archon dating formula for this period was [TEXT NOT REPRODUCIBLE IN ASCII.], and that therefore the latter should not be restored (in line 1 of fr. 1); yet, it is the one that could be made to fit. It relies on two false premises, namely, that the letters on the top line of this fragment preserve part of the dating formula [TEXT NOT REPRODUCIBLE IN ASCII.], and that fr. 1 contains the beginning of a prescript. Even if we allow both of these premises for the sake of discussion, however, the rejection of the formula [TEXT NOT REPRODUCIBLE IN ASCII.] on the basis of parallels with other lists of the period is decidedly weak. There is only one certain example of an archon dating formula in the fragments assigned to this period, namely, that used in the prescript in IG [I.sup.3] 285, for the year 421/0, [TEXT NOT REPRODUCIBLE IN ASCII.].

Meritt adduced another example, in IG [I.sup.3] 287, for the year 418/7, but there the identical formula is entirely restored. To be sure, the restoration is strong, given what is preserved of the prescript on the fragment. (81) However, the fact that the associated fragments assigned to 418/7 cannot be said to be securely dated to that year, or to any year during the Peace of Nikias, effectively eliminates any weight the restoration might otherwise have had; this leaves but one example for the entire period, that used in the prescript for 421/0. Yet, if Meritt's arguments for the date of 416/5 for frr. 1 and 2 do not hold up to scrutiny and, indeed, if it cannot be maintained with any certainty that these two fragments belong to the years of the Peace of Nikias, then the strength of even the one counter-parallel diminishes.

A further issue should be raised. Perhaps the most serious methodological weakness in Meritt's case is that his proposed restoration of the archon Arimnestos and the archon dating formula is located near the bottom of the prescript, in a position unprecedented in any other list. The two other prescripts on fragments dated to the period of the Peace of Nikias, just discussed, both have archon and verb in the beginning of the prescript. Moreover, Meritt's restored prescript in turn necessitates what he describes as a "compendious" formula ending with the phrase [TEXT NOT REPRODUCIBLE IN ASCII.] at the end of the prescript. Yet, one objection he raised to the restoration he initially considered, that with the archon name and formula appearing at the top, was that it resulted in a wordy and lengthy prescript. In any case, he has to posit a chronological development in prescripts to justify his unparalleled restoration and arrangement of the prescript, while at the same time discounting the restoration of the formula [TEXT NOT REPRODUCIBLE IN ASCII.] on the grounds that this formula was not used "during this period." These arguments overall have an arbitrary character.

Thus is the history of the list for 416/5. The editors of ATL originally thought that two of the fragments should be placed earlier on the basis of quotas similar to those for 421/0. Their movement to a less historically desirable--as they saw it--year was motivated by epigraphic grounds; but it should be remembered that one of those two fragments has long been lost, was never photographed, and cannot be examined today. Most important, the epigraphical arguments underlying the movement of all four fragments to 416/5 in some cases lack validity, and in others are valid but cannot bear the weight of certainty that Meritt accorded them. Too many possibilities have been converted to fact. It thus must be considered fundamentally uncertain whether any of these fragments--assuming that they come from the same list--should be placed in the year 416/5.

As for possible alternatives for a date, it is worth reexamining the arguments adduced in ATL against periods other than the Peace of Nikias. In ATL 1, the editors, in arguing that (old) List 35 should be dated to 421/0-419/8, before Welter's fragments came to light, excluded consideration of the period 428/7-422/1 for these fragments on what appear at first to be reasonable grounds. Notion's quota in fr. 3 (IG [I.sup.3] 289.41) is restored as [TEXT NOT REPRODUCIBLE IN ASCII.], on the basis of the list for 421/0 (IG [I.sup.3] 285, col. I.95). The editors state that Notion's tribute (2,000 drachmas, calculated from the restored quota) cannot apply to 428/7-422/1, since "we now know" that its tribute was only 100 drachmas in the 425 reassessment. (82)

On what is their certainty based? At the bottom of one fragment assigned to the 425 assessment is a nu followed by a vacat (fr. 30, line 107 of IG [I.sup.3] 71), above which are portions of place-names (Fig. 9, right). In the facsimile drawing in ATL 1, no letter is read to the left of the nu. In the drawing and on the stone, there appears to be the right curve of a letter, but it occurs at the break of the fragment. (83) In the text given, a dotted omicron appears and the name restored as [TEXT NOT REPRODUCIBLE IN ASCII.]. Meritt and West placed a sliver of a fragment containing portions of numbers (fr. 29) to the left of this fragment; at the bottom of fr. 29, along the broken left edge, appears a trace of a vertical stroke restored as H (Fig. 9, left). Thus, behind the "we now know" statement lies a nu, an uncertain omicron on one fragment, and a vertical stroke on another, nonjoining fragment.

[FIGURE 9 OMITTED]

The order in which the cities are restored on fr. 30 of the 425 assessment is not paralleled in other lists, so that the presence of Notion in that panel is not guaranteed by the restorations that precede. (84) Nevertheless, it is certainly a plausible restoration because of the unusual ending of nu, that is, a city name rather than the more common ethnic, (85) the absence of alternative cities with an ending in nu, and the relatively short length of the name. That its tribute was 100 drachmas, however, must be considered far from certain, since it cannot be taken as a fact that the fragment positioned to its left with restored H corresponding to it is correctly placed, and, even if so, other numerals may have preceded the H. (86)

In general, given the unique (from the standpoint of preservation) nature of the 425 reassessment decree and the irregularity of many assessments, neither uniformly doubled or tripled, and in some cases remaining the same as or even reduced from pre-425 assessments, (87) one should not draw any firm conclusions about Notion and its assessment in 425; the evidence that the city was assessed only 100 drachmas in 425 is extremely weak. With the exception of 428/7 (IG [I.sup.3] 283, col. III.23) in the current ordering, before and after 425 the city routinely paid 2,000 drachmas. (88) Thus, its 425 reassessment would, as currently reconstructed, constitute an enormous reduction in the city's tribute. (89)

To return to the fragments in List 39, no evidence dictates the presence of any of these nonjoining fragments in 416/5, or their association with each other. Moreover, as mentioned above, it must remain uncertain whether fr. 1 belongs to a quota list or even contains a prescript.

List 40: 4 15/4 (IG [I.sup.3] 290)

List 40 consists of five fragments containing ethnics, and preserved or partially preserved quotas for some, from an Ionic panel (col. I, frr. 1, 2) and endings of ethnics restored to comprise part of a Thracian panel (fr. 4, col. III) (Figs. 10-12). (90) They were initially dated to the period 418/7-415/4 and were combined as List 37 (418/7) in ATL 1. They were moved to their current year, 415/4, by Meritt, when he reassigned the fragments that had occupied what was then List 33 (422/1) to a list for 418/7 on the basis of the restoration Antiphon as archon in the prescript. (91) Even though, as noted above, that list was bumped back to 422/1 in ATL 2 and 3 (though subsequently reassigned again to 418/7), (92) the fragments now occupying List 40 did not also return to their former home in the year 418/7, but stayed in place in 415/4.

[FIGURES 10-12 OMITTED]

The original basis for assigning this list to the period 418/7-415/4 rested primarily on historical grounds, viz., the appearance of Miletos without reference to Leros and Teichioussa. This prompted the reasonable, but not inevitable, inference that Miletos had control of its dependencies in the syntely at the time of this list, so that the years 428/7-426/5, 425/4-422/1, and 421/0-419/8 were excluded. In fact, we know so little about Miletos's relationship to Leros and Teichioussa in this period (and in general) that hypotheses cannot carry much weight. (93) The different ways that Athens had of referring to Miletos in this period may not have been tied to political circumstances at all, but could reflect changes in the method of collection. In any case, on historical grounds, nothing ties this list specifically to the year 415/4, or even to the Peace of Nikias. Moreover, even if the terminology and the form of reference to these cities are to be explained in political terms, there is no reason why, for example, the reference in the 425 reassessment to Miletos followed by Leros and Teichioussa (IG [I.sup.3] 71, col. I.121-122) should preclude an entry with Miletos alone in a subsequent year.

Some physical and epigraphical factors also cast doubt on the association of two of these fragments. Frr. 1 and 2 are presented as having a physical join behind the surface and positioned as appearing on the same column, with a vacat of one line between them. (94) The join is possible, but the stones do not interlock, as far as I could tell on two examinations. (95) Furthermore, the lettering on the fragments differs substantially, a factor that, in turn, may reduce the probability of a join (cf. Figs. 10, 11). Clearest are the differences between alphas, epsilons, and lambdas. The horizontal cross bar on the alphas on fr. 1 is higher than those on fr. 2; two of them on fr. 1 (lines 23, 28) slope downward to the right. The epsilons on fr. 1 are upright in the stoichos with horizontal strokes measuring 0.006 m, except in line 13, where all the letters are more closely packed together. In fr. 2, the horizontal strokes have a thinner look, and measure between 0.0045 and 0.006 m, and three have a distinctly downward slope. The lambdas in fr. 1 have a more sharply angled diagonal stroke than those in fr. 2. Finally, in the combination iota, omicron, iota, the omicron in fr. 1 is regularly placed higher in the stoichos than in fr. 2. (96)

There is, in sum, no particularly strong reason for placing this list in 415/4 nor, with the exception of frr. 1 and 2, if they do indeed join, for associating the fragments at all. If frr. 1 and 2 join, the fragments have interest from the standpoint of the great variety of lettering that could occur on a single column. Most uncertain of all the extant fragments in terms of restoration and association is fr. 5 with ethnic endings. The preserved quotas on none of the fragments compel placement during the later years of the Peace of Nikias.

CONCLUSION

The above discussion underscores starkly that attempts at historical reconstruction for the period 421/0-415/4 preceding the decision ca. 413 to eliminate tribute must take as their starting point the recognition that we do not know with certainty anything about tribute assessment or collection during this important period. With one exception (fr. 1 of IG [I.sup.3] 285), none of the fragments of quota lists placed in these years necessarily belongs there. Moreover, given that each hypothesized list consists primarily of nonjoining fragments, it must remain uncertain whether fragments associated with one another are associated in every case correctly. (97)

In this examination, I have neither disproved nor aimed to disprove the current placement of the fragments, but merely attempted to demonstrate the tenuous and arbitrary nature of the grounds for assigning them to lists of this period. The editors of ATL explicitly drew on historical context as the chief criterion for assigning lists and fragments within lists. (98) This approach is not unreasonable if evidence of a particular historical setting is not itself based on undated epigraphic evidence and if the assigned lists are not in turn used to support that historical context, since this is obviously circular reasoning. With respect to the historical context of the Peace of Nikias, the editors drew on a combination of inference, assumption, and the lowest possible reading of the Hellespontine district total (in an undated assessment) to support their assignment of fragments to the years following the conclusion of peace.

If, upon closer study, a stronger case could be made than that surveyed above that the fragments currently occupying lists for the years of the Peace are correctly associated and positioned, then useful historical conclusions could be drawn: the Athenians moderated the tribute requirements of their allies for the first few years following 421, but then increased them, at least in some cases (e.g., cities on the fragment currently positioned in 417/6) in an assessment of 418/7. One might then associate this rise with discussion in the assembly already in 418 about the prospect of an expedition to Sicily, and the need therefore to boost their revenues in preparation. Such speculation is plausible, given that the Athenians had already been active in Sicily in 427, during the Archidamian War. (99)

On this reconstruction, we would then be able to maintain that the Athenians had successfully increased their tribute intake, reaching, in the case of at least some cities, the level of the 425 assessment. As mentioned at the outset, however, Thucydides (7.28.4) explicitly notes that by ca. 413 the Athenians' revenues "were perishing," which suggests that many allies had taken advantage of the Athenians' preoccupation with the West to cut back on or withhold their payments from 415 to 413. But rather than either pass another assessment decree like that in 425 to attempt to increase the tribute, or take steps to ensure better and fuller collection, the Athenians decided to abolish tribute wholesale, and substitute a maritime tax.

This scenario is possible but, as it stands, it is constructed on a house of cards. Instead, we need to focus on what is certain, and then consider plausible inferences. Moving backward in time, we know the following: (1) in ca. 413, the Athenians abolished tribute; (2) one fragment of the lists for the period 421-415 (fr. 1 of IG [I.sup.3] 285) is securely dated; (3) the Athenians demonstrably expressed dissatisfaction with tribute collection over the years preceding the decision to abolish the system; (4) in 425, the Athenians assessed tribute at an increased level; and (5) in the period between 425 and 422, the Athenians experienced significant setbacks in the war. These last three points merit further discussion.

First, inscriptions concerned with tribute illuminate a fact about Athenian majority opinion concerning tribute collection: at least at some points in the history of the arche, most Athenians either believed that they were assessing tribute, regarded as the mainstay of imperial revenue, at an insufficiently high level, or recognized that they were unable adequately to collect the assessed amount--or a combination. Incomplete and late payments of tribute, or the failure to pay at all, are reflected in the quota lists at various times, beginning early in the 440s, according to the reconstruction of the lists in ATL. (100)

Other documents are explicit in drawing attention to a perceived problem. One, the Kleinias Decree (IG [I.sup.3] 34), lays down stringent rules about collection and transport to ensure that what a city actually paid reached Athens. It is unclear whether embezzlement by those transporting the money or incomplete payments on the part of the city was at issue. Nevertheless, the decree reveals a significant problem in Athens receiving the full complement of money annually. Another document, the Kleonymos Decree (IG [I.sup.3] 68), is less ambiguous about the nature of the problem of tribute payments. It suggests that cities were failing to pay sufficiently and it orders the appointment of tribute collectors in the cities, and refers to keeping a published record of payers and defaulters on a whitened board.

Neither the Kleinias nor the Kleonymos Decree is dated; most scholars place the former in the 440s and the latter in the early 420s. (101) Fortunately for our purposes, their respective dates matter less than the fact that each bespeaks serious problems pertaining to the collection of tribute. If, however, both decrees belong in the 420s, as has been suggested, then their existence provides unequivocal evidence of the Athenians' acute concern about failures in tribute collection. Moreover, the reassessment of 425 was held in a year other than a Great Panathenaic year, the time of normal reassessment, and it evidently entailed increases in tribute. Both factors point to concern over adequate tribute collection. (102)

Finally, it is worth focusing on the period 425-422. In 425 the Athenians assessed tribute at a level higher than previous inferred assessments. (103) The decree ordering the assessment is a forceful document, and it and the assessments were inscribed on an imposing stele, set up on the Acropolis. It projected power, reflecting the Athenians' spectacular success at Sphakteria, where they forced the unprecedented surrender of about 120 Spartan hoplites (Thuc. 4.38).

One might argue that the edge that the Athenians now had in the war would have increased their ability to collect tribute. One could make a case that their strong position, combined with successful implementation of the procedures contained in the Kleonymos Decree, and perhaps also the Kleinias Decree, would have ensured vigilance and success. Indeed, the editors of ATL presumed that this was so. As mentioned above, the period following the assessment of 425 was rejected as a possible location for all but one reconstructed list on the grounds that none of the quotas reflected tribute that corresponded to a city's assessment in 425, where known, or to the severe "character" of the decree, despite the fact that it did not call for a general, sweeping dramatic increase. Some cities' assessments decreased, while others' remained consistent with previous inferred ones. (104)

The issue of tribute collection is directly tied to two considerations. First, while the assessment of 425 is a historical fact, even if there are many uncertain particulars--including the total amount of the assessment--it is only a fact of assessment, not one of collection. Yet the assumption governing the decision not to assign fragments with lower quotas to this period was that tribute collection would have corresponded to the level of assessment. (105)

This guiding assumption has resulted in a glaring anomaly in the quota record that illuminates the methodological pitfalls that can occur when it is employed in conjunction with an actual assessment decree. The period following the 425 assessment constitutes the sole continuous gap in the quota record as reconstructed chiefly because of the above assumption that, with a few exceptions, no quotas in any list equaled 1/60th of the tribute imposed in the 425 reassessment decree. It would indeed be a curious accident of recovery, however, if not a single fragment from any list for these years has been unearthed, but all other periods are generally well filled in, even if only by the tiniest, lone scrap of marble. In my view this situation should not only cause discomfiture, but also, most important, should prompt questioning of the assumptions used to position the lists.

The second consideration is the Athenians' ability to collect tribute in the years following 425. Indeed, if they had failed to collect full assessments before, how could they hope to do so later at increased levels? The historical context of those later years strongly suggests that difficulties collecting tribute in the period following 425 would have continued. In fact, the strength the Athenians possessed was fleeting. Their situation dramatically changed in 424 when the Spartan general Brasidas led a Peloponnesian army to the Thraceward district, the most valuable region in the arche to Athens because of its resources vital to the city's power, especially precious metals and timber. There he achieved the surrender of a number of Athenian subjects, and the colony of Amphipolis (Thuc. 4.102-107).

The loss of Amphipolis in particular threw the Athenians into a state of justifiable fear, for the colony had been a crucial outpost for the securing and dispatch of money and timber to Athens (Thuc. 4.108.1). Thus, the Athenians lost the ability to collect substantial revenue in the north, which would have been reflected in the quota record and in their treasury. It is also reasonable to suppose, given this context, that subjects elsewhere would have seized the opportunity to renege on their tribute payments or to submit only partial payments. The situation only continued to deteriorate, with Brasidas's continued success in achieving revolts in 423, and it caused Athens to negotiate a truce with the Spartans, which in turn was followed by the Peace of Nikias.

If some of the fragments of lists with quotas reflecting tribute at lower levels than that assessed in 425, and currently assigned to the years following 421, could be placed in the second half of the 420s, it would foster a reconstruction that would make better sense of the historical realities and attendant implications, and of the demonstrable fact of difficulty with tribute collection, than both the assumption that Athens collected the full tribute assessed in 425 and the total, curious, absence of quota records from this period. The only impediment to locating any of the lists in those years has been the invalid assumption that collection must be equated with (the presumably higher) assessment. There would be no impediment, moreover, to placing in the period following 425 the sole fragment in the lists examined here that had been restored as part of a Thracian panel, since there are no names extant of cities that revolted, with the exception of Mende. (106) It revolted in 423, but Thucydides tells us that the Athenians quickly recovered the city; its revolt was short-lived. (107)

The assessment of 425 was optimistic and unrealistic even in the Athenians' position of strength in that year, demonstrable from, for example, the presence of cities that do not appear on any extant quota list either before or after this assessment. (108) It is likely that in the following years Athens continued to experience difficulties with collection, despite forceful decisions to improve the means of collection. The 425 reassessment decree, physically imposing and intimidating, continues to impress and cow us--but perhaps more than it impressed or cowed its referents into paying up.

Indeed, it is possible, though this is mere speculation, that recognition that measures to improve collection had failed in the past led the Athenians in ca. 413 to reconsider the entire system. It is useful to recall Thucydides' comment, in connection with the Athenians' decision to abolish tribute, that their revenues "were perishing" (7.28.4), for it strongly implies that the Athenians were unable to collect what they demanded of their allies. This suggests that the allies took advantage of the Athenians' preoccupation and difficulties elsewhere, in this case in Sicily and at home in Attica, to withhold payments. It is reasonable to generalize from this that, when an ally thought it could get away with paying less than its assessment, or not paying at all, it acted on that premise. (109)

Thus, it is far from impossible that some of the fragments assigned to 421-415 could or should be relocated in the last half of the 420s--or elsewhere. Indeed, the historical context of Athens' demonstrable weakness could lend support to the placement of fragments whose quotas are lower than the assessments in 425 in the period following that assessment. In addition to the gap in the quota record for the period 425-422, there are also no fragments placed in the years following 415/4. There is good reason for this, since, as mentioned above, Thucydides states that tribute was replaced by a maritime tax in ca. 413. Meritt, however, believed that tribute was restored in 410. We have no explicit evidence for the length of time the tax was in place. (110) Thucydides provides evidence only for its imposition, not its duration. There is no reason, however, why the Athenians could not have maintained the tax and resumed the collection of tribute; this might have been a plausible decision given the extreme stress to which their navy was being put in the eastern Aegean from 412 on. If tribute was restored in the period 410-404 (or earlier than 410), then it remains another accident of recovery that not a single fragment has been unearthed that postdates 410. (111)

It is possible that Meritt and his coeditors did not assign any fragments to the years following 410 because of the insufficiency of data on these fragments from which to draw conclusions about the level of assessment and thus to match quotas with it. Yet, for the assembled quota lists from 454-425 that were unattached to any prescript necessitating their presence in a given year, a reverse approach was taken. That is, assessments were inferred and estimated from the fragments of quota lists assigned to the period that would have been covered by a presumed assessment, not because they corresponded to a known assessment and were thus placed in the period relevant to that assessment. The circularity of this approach should be obvious. (112)

The above discussion highlights a curious fact about the current arrangement of lists and also raises a larger methodological issue that applies to the reconstruction of the lists as a whole. Fragments have been positioned and associated with excessive confidence for the period 454-425, when explicit assessments of cities are unknown from any decree. By contrast, no fragments are assigned to any year following the fullest assessment record we possess, that of 425. Moreover, fragments placed in the years following the only other extant assessment (excluding the fragments situated in 410), that situated in 422/1, were assigned to those years, as we have seen, primarily on the basis of assumptions about its mild nature rather than because of actual correlations between assessed amounts and figures preserved in fragments of quota lists. (113) Most important, the notion that there are gaps of certain years falsely implies that we have lists for all others.

These larger methodological issues concerning the relationship between assessments and collection, and other assumptions guiding the assignment of lists to particular years or periods, strongly support the necessity of a thorough reexamination of the whole series of quota lists, beginning with understanding and questioning the principles, approaches, and assumptions that guided the assemblage of the lists. /////////////////////

(1.) I am grateful to Charalambos Kritzas, Director of the Epigraphic Museum (EM), Athens, for permission to examine fragments of the quota lists during 2002-2003 and for his generous assistance; to John Camp, Director of the Agora Excavations, American School of Classical Studies at Athens, for permission to examine three fragments from the Agora; and to Jan Jordan and Craig Mauzy for their assistance. I also thank Jack Kroll, Angelos Matthaiou, William K. Pritchett, Ronald Stroud, and Stephen Tracy for reading drafts of this paper at various stages, and Kroll, Stephen Lambert, Matthaiou, and Tracy for examining fragments with me in the EM. I also thank the anonymous referees for Hesperia.

(2.) Thuc. 6.26. Nikias refers to the "accumulation of money" [TEXT NOT REPRODUCIBLE IN ASCII.] in the preceding years. For discussion of the recovery of the reserve, see Samons 2000, pp. 171-248.

(3.) Various scenarios could be proposed, e.g., that tribute collection was not matching assessments; that tribute was being fully collected but that it was regarded as insufficient for what they were doing or planned to do. Thucydides mentions the decision in the context of the dual wars in which the Athenians were involved in 413, namely, fighting in Sicily and Attica, after the Spartans built a fort in Dekeleia. He comments (7.28.4) that they needed more money because "expenditures were increasing, while revenues were perishing."

(4.) Andokides also claims that the Athenians deposited "7,000 talents of coined money" into Athena's treasury, though here he is not necessarily thinking solely of tribute. The editors of ATL (3, p. 350) reasonably regard Andokides' "over 1,200 talents" as the ultimate source of Plutarch's figure.

(5.) On their dating of this assessment, however, see below.

(6.) ATL 3, pp. 346-353.

(7.) ATL 3, p. 347.

(8.) The rejection of the view of an increase in the 418/7 assessment, advocated by West (1925), depended on placing in 422/1 fragments reflecting tribute levels more in accord in some cases with the 425 assessment than the Aristeidean level to which the 422/1 assessment was thought to return. These fragments, considered below under the year 418/7, were moved back and forth between 422/1 and 418/7 precisely because they posed a problem for the view of a moderate assessment in 422/1.

(9.) West 1925. Even the editors of ATL (3, p. 348), however, noted that the treatment of the six cities may have been a concession to Sparta, along with the stipulations of autonomy and independence. Nevertheless, they still treat the Aristeidean level as the universal standard.

(10.) The assessment of 425 (IG [I.sup.3] 71) prescribed that future assessments should be held at the time of the Great Panathenaia, and contained harsh penalties for those who failed to comply (see esp. lines 26-33).

(11.) Aristophanes' Peace, produced at the Dionysia of 422/1, has also carried a great deal of weight (in American scholarship) in support of the assumption of a wholesale reduction in tribute. Consider the comment of West (1925, pp. 135-136): "Are modern historians [e.g., E. Meyer and J. Beloch] right in assuming that the high idealism and pan-Hellenic spirit of the Athenian Armistice Day, portrayed in the Peace of Aristophanes, evaporated before it found expression in an act of justice toward long-suffering allies desirous of release from their war-time burdens?" See also Meritt, Wade-Gery, and McGregor in ATL 3, p. 347: "The assessment of tribute . . . did in fact give the allies the relief promised by Aristophanes." The panhellenic warmth Aristophanes is thought to exude has been assumed by earlier scholars like West to be accompanied by a desire to help and relieve the allies. More recent commentators on Peace offer a corrective, e.g., Olson 1998, pp. 88-89 at 105-106: "A sense of common identity . . . never stopped the Greeks in any period from killing and enslaving one another. When panhellenism is put forward as a basis for collective action, therefore, it is only in the face of an external ('barbarian') threat (cf. Lys. 1128-1134) and generally in the interest of a single large power (usually Athens), whose real purpose is to obtain political hegemony over her 'ethnic' brothers."

(12.) ATL 3, p. 352.

(13.) The editors of ATL explicitly regard their restorations of epigraphical evidence as based chiefly on historical grounds (ATL 2, p. 6).

(14.) See also Olson's comments, above, n. 11.

(15.) Meiggs 1972, pp. 340-341.

(16.) Meiggs 1972, pp. 342-343.

(17.) So Samons 2000, p. 231, n. 69. I disagree, however, with his statement that "it is very hard to believe that the mason inscribing the assessment of 422/1 imitated his predecessor in 425 to the extent of reproducing his indentation." Among other considerations, the dearth of assessment decrees means that we do not know whether there was a standard practice in inscribing such documents. Moreover, the idea that the mason would imitate the format of the 425 assessment is reasonable, given that that decree was likely to have been the most imposing and memorable of assessment decrees, well worth imitating.

(18.) The presence of part of an Aktaian panel supports a date following 428/7, after the reduction of Lesbos, which had lost control of its peraia that consisted of the Aktaian cities. The Aktaian panel currently makes its epigraphic debut in 425 (IG [I.sup.3] 71, col. III.124-137).

(19.) In ATL 3, pp. 277-278, the editors argued that no tribute was collected in 449/8 (see also ATL 1, pp. 133, 175; Meritt 1972a, p. 404, n. 8), and they connected this with the Peace of Kallias; others previously had suggested 447/6 as a year without tribute collection (for bibliography, see ATL 3, p. 278, n. 16). Against the view of a missing list, see Pritchett 1964a; 1966, pp. 126-129; 1972, esp. pp. 158-159; 1995, pp. 168-171. See also Samons 2000, pp. 80-81, with n. 246, where he suggests a connection between the execution of the Hellenotamiai mentioned by Antiphon (5.69-71) and the blank space on the right lateral side of the lapis primus.

(20.) ATL 1, p. xi.

(21.) ATL 2, p. vii.

(22.) E.g., Pritchett 1963, p. 23, n. 3; 1964a, pp. 133-134; see also R. Stroud's comments in SEG XLVII 73.

(23.) An exception is current List 37 for 418/7 (IG [I.sup.3] 287).

(24.) This fragment is fundamental in dating the beginning of the practice of dedicating quota to Athena in 454/3, because in addition to an archon name, it preserves most of the formula containing the ordinal year in the series: "the thirty-fourth [year of] office in which the following cities paid the aparche to the goddess of a mna from the talent" (lines 6-7).

(25.) For photographs of all three, see ATL 1, pp. 101-103, figs. 135, 137, 138. See also Pritchett 1969, pl. 95, fig. 2; Drew-Bear and Pritchett 1973, with magnified photographs of the right edge of fr. 1, and discussion of the implications of this fragment for the Athenian calendar.

(26.) A better photograph of fr. 1 may be found in ATL 1, p. 102, fig. 136.

(27.) See Fig. 1. Fr. 2 is slightly thicker (ca. 0.13 m) than both fr. 1 (ca. 0.12 m) and fr. 3 (ca. 0.12 m). That it is thicker than fr. 1 is not an issue as it may be in the case of fr. 3, which is placed beneath fr. 2; the stele therefore would have bulged in the center. Variations in thickness of stelai, however, are not uncommon, especially when the reverse is undressed.

(28.) E.g., the funerary epigrams for Myrrhine (IG [I.sup.3] 1330) are inscribed on a stele the lower half of the obverse of which is undressed. One stele on which accounts of the Pronaos of Athena's treasury are inscribed is entirely dressed on both sides, but not inscribed all the way down on the obverse (IG [I.sup.3] 296-299, 305-308). The reverses of four associated fragments (frr. 2-5; fr. 1 is lost) of a quota list assigned to 428/7 (List 27, IG [I.sup.3] 283) raise a question similar to that posed by the fragments examined here. Fr. 2, positioned above frr. 3-5, is dressed, with an inscribed area at the left of the fragment and not at the right, while frr. 3-5, which, if correctly assigned to the same stele as fr. 1, would have been positioned farther down on the stele, also have preserved backs, and they have been evened out with a tooth chisel but not actually dressed. Note, however, that their surfaces are considerably smoother than those of frr. 2 and 3 of IG [I.sup.3] 285. I thank A. Matthaiou for confirming this by autopsy.

(29.) It may be relevant that the mason evidently was less concerned about constrictions of space for the inscription on the reverse, which does not show the same attention to spacing as that on the obverse. Letters on the reverse are spaced less closely and neatly, with less attention to aesthetics.

(30.) The editors of ATL assumed that every list after 432/1 (the last list on the lapis secundus) occupied its own stele. There are not unreasonable grounds for supposing this: fragments of lists assigned to years after 432/1 that have their backs preserved are not nearly as thick as the fragments with both front and back preserved on the lapis primus and the lapis secundus, 0.385-0.39 m (see Pritchett 1964b) and 0.34 m, respectively, compared to the thickness of frr. 1, 2, and 3 under examination here (0.12 m, 0.13 m, and 0.12 m, respectively); cf. also IG [I.sup.3] 283, frr. 2-5, with a thickness given as 0.098 m. Yet it is important to keep in mind that hardly any of the fragments have preserved backs. In general, the size of the stele and number of lists on it might have depended on the number of cities that paid tribute in a particular year, given that there would have been fluctuations, and also on whether or not the reverse was intended to be inscribed, and other unknown factors. The editors of ATL envisioned a stele that contained at least 105 lines of inscribed text; allowing for some uninscribed space, such a stele could have approached 2 m in height, with a width of ca. 0.75 m (on the basis of the width [0.39 m] of fr. 1, which appears to contain roughly the right half of the prescript). A. Matthaiou, however, having compared a number of intact stelai, thinks that the stele would have been only ca. 1.5-1.6 m in height (pers. comm.), and he may be correct; given variations in the thickness-to-height ratio, however, I suspend judgment pending further research. Geological analysis of the fragments is central to any further investigation of this issue.

(31.) Cf. Unz 1985, pp. 31-33, where the author argues, though without citing this particular inscription, that these kinds of rubrics reflect exceptions to the norm; he assumes that tribute was routinely expended in the field and only the surplus was brought to Athens, with quota comprising 1/60th of that surplus. In my view, however, this argument involves too much special pleading to be convincing.

(32.) In either case, whether it pertains to 421/0 or, e.g., the following year, it still provides valuable evidence of military activity in this region.

(33.) ATL 1, p. 104.

(34.) I reserve discussion of these fragments for the examination of all the fragments currently assigned to the year 416/5; see below, pp. 480-487.

(35.) ATL 2, p. 35. Cf. also Meritt 1941, p. 15: "Only lines 1-7 of A.T.L., I, List 35 remain as a possible candidate, and this is uncertain."

(36.) This fragment was first published by Broneer (1935, pp. 157-158), along with an excellent photograph (p. 157, fig. 46).

(37.) As arranged in ATL spatially and chronologically (on the basis of their reconstructed layout of IG [I.sup.3] 71, the reassessment decree of 425), only in that year does the Island panel begin to precede the Ionic panel.

(38.) ATL 1, p. 201.

(39.) Meritt 1941, p. 3.

(40.) I explore the historical context of this period below.

(41.) Meritt 1941, p. 15.

(42.) On this issue, see below, pp. 494-495.

(43.) In publishing fr. 5, Meritt (1972b, p. 418) commented that "this fragment makes a join beneath the earlier known fragment 1 and to the left of the earlier known fragment 3 of List 37 of 418/7 B.C." The joins are possible, especially in the case of frr. 1 and 3, with fr. 1 containing the beginning of the Hellespontine district heading and fr. 3 continuing it with the correct amount of space in between for the one missing letter ([SIGMA]). This join is impossible to confirm at this time, however, because fr. 1 (Agora I 4809), on display in the Agora Museum, is bolted to a base.

(44.) On fr. 2, the second to last line contains the letters A[DELTA]E, and the bottom line, AI[DELTA]E. It is clear that these letters come from the end of a prescript of a quota list since the Island panel begins below it (in what would have been col. I). Fr. 3 preserves most of the phrase [TEXT NOT REPRODUCIBLE IN ASCII.].

(45.) In addition to the end of a prescript, fr. 3 contains part of the Hellespontine panel with [PI]ONTIO preserved at the beginning; fr. 5 is part of the Hellespontine panel with the beginnings of 14 ethnics and ends of numerals in four cases; fr. 4 (Agora I 4809b) is a tiny fragment with little usable information. It has parts of four lines with only four letters fully preserved and traces of others allowing for more than one possibility. The first line has the bottom horizontal stroke of a letter, printed as a certain epsilon. Line 2 has XE, line 3, an alpha followed by a tiny left vertical upper tip of a letter; line 4 has a kappa.

(46.) Fr. 3, as mentioned above, n. 44, preserves most of the final phrase of the prescript, [TEXT NOT REPRODUCIBLE IN ASCII.] (line 7). While the preserved bit of the prescript in fr. 2 does not by itself allow us to determine the number of letter spaces between it and the left edge, the left half of the island heading and end of one numeral permit an estimated line length in the prescript of 51 letters. The precise line length is uncertain, however, and one or two vacant spaces must be assigned at the end of line 2 (and one at the end of line 1), depending on whether a name of seven or eight letters is restored.

(47.) See also Meritt 1939, pp. 56-57. If frr. 1, 2, and 3 are correctly associated with one another, they yield sufficient portions, especially in the last line, of the prescript to estimate the line length, which in turn can be used to estimate the number of letters available for the archon name in line 2.

(48.) Col. II.10. The quota XX is on fr. 1 (Agora I 4809) and the ethnic [TEXT NOT REPRODUCIBLE IN ASCII.] is on fr. 3 (EM 12798). This inference is valid only if the two fragments do in fact join; see above, n. 43, and below.

(49.) Meritt 1939, pp. 54-59, esp. p. 58.

(50.) Meritt 1941, p. 10.

(51.) Meiggs 1950, p. 64.

(52.) Meiggs 1950, p. 64.

(53.) ATL 3, p. 352. They also concluded (pp. 351-352), after the final relocation of lists for the period 418/7-415/4, that the increase in tribute in 418/7 was less sharp than they had previously supposed.

(54.) Meritt 1972b, p. 420.

(55.) Didymoteichitai and Daunioteichitai, lines 19 and 20, respectively.

(56.) Meritt 1939, p. 55.

(57.) Meritt 1939, p. 55.

(58.) Meritt 1936, pp. 386-389; see also ATL 3, pp. 91-92. Mattingly (1996, pp. 158-160, 205-208) places the fragments in 418 (and argues against the resumption of tribute). Samons (2000, p. 251) considers 414 possible; for discussion of these dates, see also Meiggs 1972, pp. 438-439. See also below, pp. 494-495.

(59.) Samons (2000, p. 254) suggests this possibility. For the likelihood that the eikoste of ca. 413 was unsuccessful in securing for the Athenians greater revenue or, alternatively, for the possibility that the Athenians collected both the tax and tribute, see Kallet 2001, pp. 222-225. The year 412 was one of crisis for the Athenians, according to Thucydides (8.15.1), for in the summer of 412 the revolt of Chios made them panic over their financial resources, and this collective state of mind led them to liberate the "Iron Reserve" of 1,000 talents, locked away since the beginning of the Peloponnesian War and not intended for use except in the event of an enemy attack on Athens by sea (Thuc. 2.24.1). It is by no means impossible that their panic also led them to reimpose tribute at the same time, with the first collection to be made in spring 412/1, though in that case one might wonder why Thucydides would have chosen to be silent about it. One explanation for his silence might be that he had already registered his view that it was mistaken to think that eliminating tribute and imposing a maritime tax would increase the revenue reaching Athens. For an analysis of Thucydides' narrative presentation of this decision, and of that to use the "Iron Reserve," see Kallet 2001, pp. 136-146, 246-250. I am not endorsing 412/1, but rather using the possibility to test its viability as a date compared to 418/7 when one adduces historical arguments.

(60.) I thank Stephen Tracy for examining the lettering on the EM fragments with me. In fr. 2 (EM 6784), the right vertical of the nu appears both angled and straight, but the base of the right stroke is consistently higher than that of the left (Fig. 3). In fr. 3 (EM 12798), the nus are uniform and straight, and the right stroke extends farther down (Fig. 4).

(61.) Unfortunately, no nus are preserved on the portion of fr. 2 containing part of the prescript.

(62.) The sigmas on fr. 1 are larger and less sharply angled than those on fr. 3. The sigmas on fr. 2 appear to be smaller, but this may be due to the condition of the stone. The sigmas on frr. 3 and 5 are compatible with each other.

(63.) It is one of those frustrating accidents of preservation that the sigma that would have been inscribed for the Hellespontine panel heading is not extant: fr. 1 preserves [TEXT NOT REPRODUCIBLE IN ASCII.] and fr. 3, [PI]ONTIO.

(64.) For a photograph, see ATL 1, p. 105, fig. 141.

(65.) ATL 1, p. 202.

(66.) Meritt 1941, p. 15 (my emphasis).

(67.) See pp. 474-475, above.

(68.) ATL 1, p. 201: "The evidence proves convincingly enough that the list cannot be dated before 421/0; and since we have in List 34 the record for 421/0, the present inscription falls in 420/19 or later. It is true that the scanty evidence now at our disposal does not allow us to exclude the document from the period 418/7-415/4. Yet in view of the parallels between 34 and 35 we have deemed it best at this time to assign List 35 to 420/19 or 419/8."

(69.) Welter 1939; Meritt 1941 (with photographs, figs. 1, 2; see also ATL 2, p. 37, figs. 1, 2). The two fragments were excavated from the Nike bastion, but Welter, although he thought that they were both pieces of quota lists, dissociated them temporally. It is impossible to confirm that frr. 3 and 4 resemble each other, since the latter is missing. Meritt's association of EM 6650 and the two new fragments was based on resemblance in letter shapes and dimensions, measurements in line spacing, and flaking of the stone. For a photograph of EM 6650, see ATL 1, p. 104, fig. 140. Fr. 1 of IG [I.sup.3] 286 (EM 12789), examined above, was regarded as sufficiently different not to warrant moving it along with EM 6650.

(70.) [TEXT NOT REPRODUCIBLE IN ASCII.], lines 5-6; Meritt 1941, p. 2.

(71.) Meritt 1941, p. 5 (my emphasis).

(72.) Meritt 1941, p. 5.

(73.) Meritt 1941, p. 5.

(74.) Meritt 1941, pp. 5-6; this reading is identical to that printed in IG [I.sup.3], except that in the latter publication, in line 1 the epsilon is no longer dotted, and in line 7 the initial tau is dotted. It is unclear whether these changes were intentional.

(75.) The stone is so badly worn that it is nearly illegible (Fig. 7). I made a number of attempts to read it, under various lighting conditions, holding the fragment at different angles, and using water and magnification, all with limited success. A. Matthaiou, examining it with me on one occasion, was able to hold it at an angle under particular lighting conditions so that many letters were visible, though the top line much less so than the other lines. The fragment could have weathered since Oliver examined it; but even Meritt refers to the extent of erosion on the surface, a problem that the photographs published in Meritt 1941 and ATL 2 reveal was present at that time. It is worth noting, however, that there may be a bit of preserved back at the lower right of fr. 2, and if so, then that fragment is thinner than fr. 1, the back of which is not preserved. Since fr. 2 is positioned lower than fr. 1, this factor may constitute a possible objection to the association, though it cannot be pressed.

(76.) The top of the fragment is rounded at the back, which is suggestive of a top edge of a stele. EM 13049 (fr. 2) has a similarly rounded or beveled edge, but fr. 1 clearly has no preserved edges. Moreover, it is not clear that there is an alpha in the first line, as read by Welter (1939). There is only part of a diagonal stroke, the angle of which seems to be too sharp, at nearly 45 degrees, for an alpha, and instead suggests a chi. It should be said, however, that this cannot be regarded as certain, again, because of the condition of the stone.

(77.) Moreover, under certain conditions, I thought I might be seeing a right vertical stroke compatible with pi. But see also below.

(78.) Meritt (1941, p. 5), however, had rightly dotted both epsilon and iota in line 1, and the underdots remained in ATL 2, p. 36, so the absence in IG of the underdot for the epsilon may be an inadvertent misprint.

(79.) A. Matthaiou (pers. comm., examining it with me) arrived at the same reading independently.

(80.) As mentioned above, it is uncertain whether a top edge is preserved. If the fragment does preserve part of a prescript and the letters on the first preserved line are E[SIGMA]AR, Matthaiou (pers. comm.) suggests the appealing restoration [TEXT NOT REPRODUCIBLE IN ASCII.].

(81.) Fr. 1 (Agora I 4809). On the first line of the fragment, most of the name Aphidnaios appears, followed by a certain pi, rho, omicron, and a dotted tau, securing the restoration [TEXT NOT REPRODUCIBLE IN ASCII.]. The fragment contains the beginning of a column of entries at its lower right, with space to the left. The inscription is restored as containing two columns, and the line length of the prescript restored accordingly. It is possible, however, that there were three columns, not two.

(82.) ATL 1, p. 198.

(83.) In Meritt's squeeze the possibility of reading anything is vitiated by the drawing in black marker of the right curve of an omicron. The setting of the fragment in plaster prevents complete inspection. That there is a curved letter stroke seems certain from the brownish coloration of the stroke compared to that of preserved letters on the fragment. Epigraphically, a letter other than omicron is possible, e.g., theta or rho, either of which would contain greater theoretical probability because the position of the curve in the stoichos is lower and farther to the left than for omicrons elsewhere (though this last point presumes that all comparanda on the reconstructed stele are correctly assigned to this inscription). Given the nu following it, however, orthographically an omicron is dictated.

(84.) As restored in IG [I.sup.3] 71, the order of the cities (or ethnics) immediately above the restored Notion is Chalkeatai, Chioi Kares, Kyrbissos. In IG [I.sup.3] 279 the order (restored) is Pyrnioi, Naxiatai, Hairaioi, Kolophonioi, Noties, and, in 280, Marethesioi, Koioi, Hairaioi, Kolophonioi, Noties. One should note, at the same time, that ordering can differ widely on fragments of lists.

(85.) While [TEXT NOT REPRODUCIBLE IN ASCII.] appears in IG [I.sup.3] 285, most often the ethnic [TEXT NOT REPRODUCIBLE IN ASCII.] is used, as in, e.g., IG [I.sup.3] 259-264, 270, 279, and 280 (and the ethnic is the basis for restoration elsewhere).

(86.) As reconstructed and restored, the left edge of fr. 29 is considered equivalent to the left edge of the column.

(87.) As the lists and the 425 reassessment decree are currently reconstructed, Pedases paid 2 T in 453/2 and 451/0 (quota partly restored), 1 T in 447/6, and was assessed 3,000 dr. in 425; Karuandes paid 1,000 dr. in 452/1 (quota almost entirely restored), went down to 500 in 450/49, and was then assessed again at 1,000 dr. in 425; Kyrbissos paid 2,000 dr. in 454/3, 452/1, 447/6, 446/5 (mostly restored), and 445/4, and was assessed the same in 425; Chalkeatai paid 3,000 dr. in 450/49, and 2,000 dr. in 445/4, 443/2, 442/1, and 432/1 (though it again appears with the rubric [TEXT NOT REPRODUCIBLE IN ASCII.]), and was assessed 2,000 dr. in 425; Broukountios paid 500 dr. in 448/7 and 447/6, and was assessed the same in 425. Karystos paid 7.3 T in 451/0, and 5 T in 450/49, 448/7, 447/6, 443/2, and was assessed that figure in 425. Finally, Notion paid 2,000 dr. in the lists preceding 425, and was assessed 100 dr. in 425.

(88.) IG [I.sup.3] 259, col. III.22; IG [I.sup.3] 260, col. VI.11; IG [I.sup.3] 261, col. V.11; IG [I.sup.3] 262, col. IV.6; IG [I.sup.3] 263, col. II.31; IG [I.sup.3] 264, col. II.29 (end restored); IG [I.sup.3] 270, col. I.8; IG [I.sup.3] 272, col. I.24 (one numeral restored); IG [I.sup.3] 280, col. I.39 (first two numerals restored, third dotted, remaining five preserved); IG [I.sup.3] 285, col. I.95.

(89.) One could, of course, explain it. In the exceptional year, 428/7, in the current assignment of fragments to this list and this list to this year, Notion paid only 100 drachmas. If both this fragment and the list to which it belongs are correctly dated to this year, then Notion would probably have been taking advantage of the revolt of Mytilene that occurred in that year. Thucydides tells us (3.34.4) that "later" (Isteron), the Athenians colonized the site (i.e., Notion). See Hornblower 1991 at 3.34.4; see also Malkin 1987, pp. 101-102, n. 69, 255-256. Malkin argues that Athenians would have been "organizers" rather than colonists themselves; if so, Notion's financial obligations to Athens may have differed from what they would have been if the city had become a settlement of Athenian citizens. It is possible, then, that a temporary (or permanent) reduction in tribute accompanied a political settlement.

(90.) EM 6653, 6756, 120, 6702, 6757. EM 120 is lost. For photographs of all fragments, see ATL 1, pp. 105-106. Fr. 3, containing numerals, is positioned on col. I. Fr. 5 (Fig. 12), in the position of a col. III, is extremely fragmentary, containing only ends of four ethnics at the beginning and the letter gamma (or alpha or delta) followed by IO[SIGMA]ENTA five lines below. In IG [I.sup.3], a certain gamma is printed, despite the admission in ATL 1, p. 106, that alpha or delta were also possible. The Thracian ethnics, Pleumes, Maroneies, and Galephsioi, are heavily restored--only the endings [TEXT NOT REPRODUCIBLE IN ASCII.], [TEXT NOT REPRODUCIBLE IN ASCII.], and [TEXT NOT REPRODUCIBLE IN ASCII.] are printed as certain. In the case of [TEXT NOT REPRODUCIBLE IN ASCII.], only the bottom tip of a letter stroke preceding epsilon is preserved; Matthaiou, examining it with me, thinks that it is too vertical to be part of a mu. (It is impossible to check the accuracy of the reading [SIGMA]I since there is a deep chip in the stone there.) Since this is a nonjoining fragment, it is not inevitable that these names were part of a Thracian panel. It is especially difficult to understand the rationale behind the restoration of these particular cities. The ending of restored Pleumes extends as far as that of restored Maroneies (see Fig. 10), while the last letter of restored Galephsioi falls under the iota of Maroneies and between the restored mu and epsilon of Pleumes. Since the ethnics would have been aligned on the left edge of the column, it is in fact an impossible combination, since, even if not stoichedon, the beginnings of the names would not come close to aligning. Most damaging to the restoration is Matthaiou's reading of a trace of the bottom of a vertical stroke before the epsilon compatible with iota (or tau, though orthographically that is less likely). He saw no trace of an angled stroke. Moreover, he tentatively thought that there might be a trace of a bottom horizontal stroke in the letter space preceding it, in which case the ending of the ethnic could be [TEXT NOT REPRODUCIBLE IN ASCII.].

(91.) See above, p. 477.

(92.) See above, p. 477.

(93.) Cf. Gorman 2001, p. 49, where it is noted that Miletos was "firmly in possession" of Leros by 454, when the island is listed on the first quota list (listing Teichioussa as well, IG [I.sup.3] 259, col. VI.19-22).

(94.) ATL 1, p. 105.

(95.) I thank Stephen Tracy and Jack Kroll for examining the fragments with me on separate occasions; both questioned the certainty of a join.

(96.) The phis on fr. 2 have saucerlike, rather squat circles that are inscribed below the middle of the vertical stroke and tilt slightly downward from left to right. The one preserved phi on fr. 1 is difficult to read, but, when moistened on one occasion, it appeared more rounded and centered along the vertical stroke, without a tilt. On a separate occasion, however, I was less confident that a substantial difference existed, especially because of the condition of the stone's surface where phi appears. I do not deal with the lettering on fr. 4 (EM 6702) only because this fragment is positioned in a different column (III, as is fr. 5), and therefore any differences could be accounted for by positing another mason at work. Overall, it resembles fr. 1 more than fr. 2, though not consistently.

(97.) Regarding the issue of associating fragments, and the need to take into account geological as much as epigraphical considerations, see the discussion in Herz and Pritchett 1953 (especially for methodological issues); and Pritchett 1953, pp. 235-236 (for application of the methodology to the Attic Stelai).

(98.) See, e.g., ATL 2, p. 6.

(99.) See Thuc. 3.86 for the initial dispatch of ships. The Athenians sent out additional ships over the next two years. The Egesta Decree, dated to 418/7, would further support this reconstruction. See Chambers, Galluci, and Spanos 1990. The bibliography on their redating is now substantial. See Mattingly 1999, p. 117, n. 1, for dissenters. See also Matthaiou 2004 (in support).

(100.) See ATL 3, pp. 67-69, 80-89; ML 39 for discussion. For the rubrics in the quota lists, see Lepper 1962.

(101.) For discussion see ML 46 and 68; Mattingly (1996, pp. 11, 285-286), however, argues that Kleonymos's decree precedes Kleinias's, and he places them both in the 420s. See also Fornara and Samons 1991, pp. 179-181; Samons 2000, pp. 184-193.

(102.) Thucydides' reports of "money-collecting" expeditions to Karia and Lykia in 429 (2.69) and the north Aegean and Hellespont in 425 (4.50; 4.75) may be relevant as well. These are almost certainly to be viewed as simple extortion of additional moneys, either from cities who were paying tribute, or from those who were not (or both), and the practice may suggest an inability to collect the full complement of regular revenue. For discussion of these ships, see most recently Samons 2000, pp. 194-195; also, Kallet-Marx 1993, pp. 136-137, 160-164, against the view put forth in, e.g., ATL 3, pp. 69-70, 89, that these ships were sent out in conjunction with tribute reassessments. Cf. also Gomme (1956, pp. 202-203), who connects them with regular collections of tribute in various capacities; Schuller (1974, p. 56) also associates them with tribute collection, while Meiggs (1972, p. 254) expresses doubts.

(103.) IG [I.sup.3] 71. As bad luck would have it, the first figure of the grand total is not preserved. A restoration of ca. 960-1,000 T is possible, as is ca. 1,460-1,500 T. Meritt and West (1934, pp. 78-90) argued that the higher figure was the correct one, on the basis of two partially preserved district totals that showed a sharp increase (doubling or trebling) from the (inferred) previous assessment. If the remaining district totals were similarly increased to twice to three times what was assessed previously, the grand total would reach over 1,460 talents. It is important to recognize, however, how much is uncertain about this document from the standpoint of reconstruction and restoration; the lower figure must remain a possibility, even if a less likely one. Cf. also ML 69; Kallet-Marx 1993, pp. 164-170.

(104.) See n. 87 above.

(105.) A typical illustration is Meritt and McGregor's (1967, p. 86) rejection of 422/1 as a candidate for a fragment of a quota list containing the ethnic Siphnioi. Decisive evidence, as they saw it, against placing Siphnos in the period between the 425 assessment and the Peace of Nikias was that the quota (unpreserved, but with a smooth surface where a numeral would have had to have been) would not have been equivalent to 1/60th of the Siphnians' assessment of 900 dr.

(106.) IG [I.sup.3] 290, fr. 4, line 10.

(107.) For the revolt, see Thuc. 4.123; for the recovery, 4.129. Galepsos, one city that revolted, is, as noted above, almost entirely restored on fr. 5 of the same reconstructed list. See n. 90 above for the problem with this restoration, and the possibility that this fragment is not even from a Thracian panel.

(108.) E.g., Melos and Thera (IG [I.sup.3] 71, col. I.65, 68).

(109.) It is important to bear in mind, however, that certain allies were responsible for submitting other forms of revenue besides tribute, e.g., rents, and that these could have been withheld or insufficiently paid.

(110.) For discussion of this point, see Samons 2000, pp. 250-254; Kallet 2001, pp. 222-225.

(111.) In that case, Thucydides' silence on so many issues that we regard as important should not obtrude unduly, especially since his work breaks off in 411. That he says nothing about a reimposition of tribute at the time that he notes its elimination need not mean that it did not occur, whether or not the eikoste remained in place, at least on the books.

(112.) Related to this, however, I recognize that the thickness of fragments is a critical factor in the assignment of fragments to, most prominently, the Lapis Primus.

(113.) As mentioned above, p. 468, a record of only a handful of cities with assessments is preserved.

REFERENCES

ATL = B. D. Meritt, H. T. Wade-Gery, and M. F. McGregor, The Athenian Tribute Lists 1 (1939); 2 (1949); 3 (1950), Cambridge, Mass.

Broneer, O. 1935. "Excavations on the North Slope of the Acropolis in Athens, 1933-1934," Hesperia 4, pp. 109-188.

Chambers, M. H., R. Galluci, and P. Spanos. 1990. "Athens' Alliance with Egesta in the Year of Antiphon," ZPE 83, pp. 38-63.

Drew-Bear, T., and W. K. Pritchett. 1973. "IG I2, 220: Affidavits (Plates 1, 2)," REG 86, pp. 35-44.

Fornara, C. W., and L. J. Samons II. 1991. Athens from Cleisthenes to Pericles, Berkeley.

Gomme, A. W. 1956. A Historical Commentary on Thucydides 2, Oxford.

Gorman, V. 2001. Miletos: The Ornament of Ionia, Ann Arbor.

Herz, N., and W. K. Pritchett. 1953. "Marble in Attic Epigraphy," AJA 57, pp. 71-83.

Hornblower, S. 1991. A Commentary on Thucydides 1, Oxford.

Kallet, L. 2001. Money and the Corrosion of Power in Thucydides: The Sicilian Expedition and Its Aftermath, Berkeley.

Kallet-Marx, L. 1993. Money, Expense, and Naval Power in Thucydides' History 1-5.24, Berkeley.

Lepper, F. A. 1962. "Some Rubrics in the Athenian Quota Lists," JHS 82, pp. 25-55.

Malkin, I. 1987. Religion and Colonization in Ancient Greece, Leiden.

Matthaiou, A. P. 2004. [TEXT NOT REPRODUCIBLE IN ASCII.] IG [I.sup.3] 11," in [TEXT NOT REPRODUCIBLE IN ASCII.] Adolph Wilhelm (1864-1950), ed. A. P. Matthaiou, Athens, pp. 99-124.

Mattingly, H. 1996. The Athenian Empire Restored: Epigraphic and Historical Studies, Ann Arbor.

--. 1999. "What Are the Right Dating Criteria for Fifth-Century Attic Texts?" ZPE 126, pp. 117-122.

Meiggs, R. 1950. "The Athenian Tribute Lists," CR 64, pp. 53-65.

--. 1972. The Athenian Empire, Oxford.

Meritt, B. D. 1936. "Greek Inscriptions," Hesperia 5, pp. 355-441.

--. 1939. "Greek Inscriptions (14-27)," Hesperia 8, pp. 48-90.

--. 1941. "New Fragments of the Tribute Lists," AJP 62, pp. 1-15.

--. 1972a. "The Tribute Quota List of 454/3," Hesperia 41, pp. 403-417.

--. 1972b. "Two New Fragments of the Tribute Lists," Hesperia 41, pp. 418-421.

Meritt, B. D., and M. F. McGregor. 1967. "The Athenian Quota List of 421/0 B.C.," Phoenix 21, pp. 85-91.

Meritt, B. D., and A. B. West. 1934. The Athenian Assessment of 425 B.C., Ann Arbor.

ML = R. Meiggs and D. Lewis, eds., A Selection of Greek Historical Inscriptions to the End of the Fifth Century B.C., rev. ed., Oxford 1988.

Olson, S. D., ed. 1998. Aristophanes' Peace, Oxford.

Pritchett, W. K. 1953. "The Attic Stelai, Part I," Hesperia 22, pp. 225-311.

--. 1963. "The Three-Barred Sigma at Kos," BCH 87, pp. 20-23.

--. 1964a. "The Height of the Lapis Primus," Historia 13, pp. 129-134.

--. 1964b. "The Width of the Lapis Primus: A Correction," CP 59, p. 272.

--. 1966. "The Top of the Lapis Primus," GRBS 7, pp. 123-129.

--. 1969. "Two Illustrated Epigraphical Notes," AJA 73, pp. 367-370.

--. 1972. "Lucubrationes epigraphicae," CSCA 5, pp. 153-181.

--. 1995. Thucydides' Pentekontaetia and Other Essays, Amsterdam.

Samons, L. J., II. 2000. Empire of the Owl: Athenian Imperial Finance (Historia Einzelschriften 142), Stuttgart.

Schuller, W. 1974. Die Herrschaft der Athener im Ersten Attischen Seebund, Berlin.

Unz, R. K. 1985. "The Surplus of the Athenian Phoros," GRBS 26, pp. 21-42.

Welter, G. 1939. "Vom Nikepyrgos," JdI 54, Beiblatt, pp. 2-22.

West, A. B. 1925. "Aristidean Tribute in the Assessment of 421 B.C.," AJA 29, pp. 135-151.

Lisa Kallet

UNIVERSITY OF TEXAS

DEPARTMENT OF CLASSICS

1 UNIVERSITY STATION C3400

AUSTIN, TEXAS 78712

kallet@mail.utexas.edu
COPYRIGHT 2004 The American School of Classical Studies at Athens (ASCSA)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kallet, Lisa
Publication:Hesperia
Geographic Code:4EUGR
Date:Sep 22, 2004
Words:19186
Next Article:Strabo 10.2.4 and the synoecism of "Newer" Pleuron.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters