The sound world of Art Tatum.

When a French journalist asked writer Andre Gide in 1905 whom he considered to be France's greatest poet, Gide responded "Victor Hugo, helas!" (quoted in Richardson 1976, 294). It became a celebrated remark and must rank as one of the most concise yet pregnant summaries of an ambivalent response to an artist's reputation ever uttered. Beginning--as one might imagine hearing Gide speak it--with sharply rising anticipation, it dissolves instantly into disappointment. Although none has expressed it quite so succinctly, many jazz writers have felt much the same about Art Tatum. A not-uncommon judgment is that he had an unrivaled, exciting technique but lacked the creative imagination to put that technique to maximum use. "A superficial bravura technique a la Paganini," one writer called it, adding for good measure, "and what interest does that merit?" (Wiedemann 1955, 28).

Tatum's admirers have been legion, of course, evidence of which can be found in a range of sources, from comments by musicians in interviews to letters in magazines, from articles in the jazz press to entries in reference books. Almost without exception, they emphasize in particular the originality of his style and the marriage of technique to harmonic innovation, but it is not unusual also to find assessments that amount to a description of the complete musician. In The New Grove Dictionary of Jazz, for example, Felicity Howlett and J. Bradford Robinson (1988, 519) write that "Tatum's technical abilities, lightness of touch, and control of a full range of the instrument were unprecedented.... [H]e had an unerring sense of rhythm and swing, a seemingly unlimited capacity to expand and enrich a melody, and a profound and continually evolving grasp of substitute harmonies."

One of the problems with which jazz historiography has been faced is Tatum's evident lack of fit into established narratives and agreed values. Among musicians (where reverence for him was almost unanimous), some solved the problem by ceasing to consider him as a jazz musician. As fellow pianist and contemporary Teddy Wilson told an interviewer after Tatum's death, "Back in the old days, we put Tatum in a special category and did not discuss him as a jazz pianist--he was in a category by himself, and we then talked about the others; those who played in bands" (quoted in Young 1963, 23). In a similar vein, more recently, critic Gary Giddins (1998, 439) entitled an essay on Tatum, "Art Tatum (Sui Generis)."

Taken as a whole, jazz historiography seems to have resigned itself to a bemused ambivalence in regard to Tatum and to have postponed resolving the issue by consigning him to the special kind of marginality reserved for talented non sequiturs. As a consequence, not only is Tatum underrepresented in jazz criticism but his presence in jazz historiography seems largely to prompt no particular effort in historians beyond descriptive writing designed to summarize his pianistic approach. Had his biography been more dramatic, his behavior more unpredictable, his personality more charismatic--had he been one of those whom jazz historiography has seen as "spectacularly socially dysfunctional practitioners available for romanticisation" (Johnson 2002)wit is possible, given the way that these things work, that his music would be thought to have signification on other levels. Although evidently fiercely proud of his skills, Tatum was an easy-going generous spirit, whose lifestyle, tuned though it was to the demands of late-night (often all-night) music making, appeared to have absorbed little of its more "colorful" side beyond a fatal taste for drink.

The combination of disagreement over Tatum's status and the relative marginality of his position in much jazz historiography suggests that we are in the presence of processes of generic canonization. Gary Tomlinson (1991, 245) has described the jazz canon: "a strategy for exclusion, a closed and elite collection of `classic' works that together define what is and isn't jazz." The canon has been established through the intervention of what Tomlinson calls an "internalist ideology," which accords absolute priority to musical features and in the process distances the music, as he puts it, "from the complex and largely extramusical negotiations that made it and sustain it" (247-248).

But there are alternatives. Internalism and canons, in Tomlinson's view, constitute "a narrowly based value judgement that cannot do justice to the complex dialogues of self and other in which culture is created" (248-249). And it is in this spirit that I want to suggest not only that Tatum stands in need of reconsideration but that such reconsideration can help us reassess our critical processes. One can make a case for suggesting that Tatum deserves renewed attention on technical, formalist grounds alone. There is a need for further close examination of his technique, free from prejudice or unbridled enthusiasm, building on the pioneering work of Felicity Howlett (1983). But that should not--cannot--be all that there is to say. The very fact that Tatum's music was (and to a large extent still is) in tension with the dominant critical paradigms suggests that the "dialogues of self and other" that underlie cultural production are indeed complex in this case and deserve some unraveling.

On the face of it, Tatum may appear a perverse choice of a musician around whom to examine such a subject. Where some recent jazz scholarship, under the influence of ethnomusicology, has emphasized the interactive nature of jazz, the dominance of solo performance in Tatum's career tends to leave him saddled with descriptions such as Gunther Schuller's (1989, 477) that refer to the "solitary nature of his art" and that see him as "artistically a loner." That he was so revered by fellow musicians may only make him less malleable to the purposes of cultural analysis; a "musicians' musician," as Max Jones (1956, 8) once described him, may sound unpromising territory on which to uncover the dialogical drama in the interplay of cultures. But the point is not to test a method on an individual; rather, it is the reverse. The aim is not to replace one set of dominant paradigms with another but to attempt to understand something of the specificity of an individual self, occupying its own unique place and time and articulating itself--performing, if you will--in and through sets of dialogical relationships.

Such activity rarely takes place on an even playing field and almost invariably involves processes of negotiation with hegemonic discourses. It will be necessary, therefore, to look again at some of the elements that have made Art Tatum controversial. But first we need to locate him.

Locating Tatum

First, a brief but essential look at his early biography. Art Tatum was born in Toledo, Ohio, in 1909. He was blind in one eye and had limited sight in the other. It is hard at this distance to ascertain the extent to which, as a black youngster, Tatum would have had a significantly different experience of music from that of white youngsters in Toledo. The black population of Toledo was small, between 1 and 2 percent around 1910, and the Tatums lived in a relatively unsegregated neighborhood. For all children in that neighborhood, music could be encountered live in a designated number of public places, some open (especially the street with its marching bands, perhaps street musicians), some in framed spaces, chief among which would be the church and the school, perhaps also the theater or the tent show. It was also encountered in the home, both live and--as yet in small measure--in technically reproduced form. In the 1910s, the latter meant a phonograph for some, but more commonly a piano-roll player. The Tatums had a piano at home, and Art began to experiment on it early, but there is no evidence of their being a particularly musical family. At some point, Tatum had a black piano teacher, Overton G. Rainey, who may have given Tatum the basis of piano technique and a familiarity with the popular classical repertoire. It is unlikely that Rainey encouraged his pupils in African-American idioms (Lester 1994, 37-38). Tatum also attended the Toledo School of Music, but for the most part he seems to have been self-taught.

By the mid-1920s, Tatum had become celebrated locally as a gifted and versatile pianist. According to his biographer James Lester, he was in regular demand as a provider of music for social occasions, playing "probably one-steps, two-steps, cake-walks, and rags that were still so popular then" (42)--popular across the spectrum of society in which a young black pianist could move in Toledo. Despite the unsegregated nature of the Tatums' domestic environment, the world of public entertainment was a different matter. When he began to enter the twilight world of clubs and bars in his later teenage years, Tatum's experience was unusual in one sense, in that his family, although priding themselves on their respectability, seem to have raised no particular objection (19).

In the midst of this young black pianist's expanding sound world, jazz emerged to play a crucial role. By the early 1920s, jazz had become the first music associated with African-American culture to achieve a widespread penetration into homes and communities via the new medium of recording. Tatum taught himself to copy recordings by New York-based pianists James P. Johnson and Fats Waller, among others, and by 1926 was playing professionally in Toledo clubs where jazz was performed. By 1929, he was performing on radio in the area. In 1932, he left Toledo, hired by touring singer Adelaide Hall as one of her pianists, and the following year made his first solo recordings for Brunswick in New York. These moves provided the first--and quite dramatic--steps from a local and regional celebrity to a national reputation that would sustain him, albeit with ups and downs, until the end of his life. Art Tatum died in Los Angeles in 1956.

Many people have commented on the apparent gap between the musically undistinguished surroundings of Tatum's early years and the scale of his technical achievement. Teddy Wilson, for example, declared that he had "never been able to trace the influence in Tatum--where and how he evolved that way of playing in Toledo, Ohio." Noting that Tatum "never played with a rhythm section before he came to New York," Wilson remarked in particular that Tatum "was improvis[ing] those harmonies ... long before he came East. No other pianist had, even remotely, that conception of playing" (quoted in Young 1963, 23).

Clearly, Tatum's was a special talent that no amount of contextualization can explain. But we can attempt to track the roots of his perspective across the interactive flow of music and musicians that peopled his world. Although it may be possible to argue a case for placing this perspective entirely, or almost entirely, within African-American culture and to portray Tatum as one particular product of forces at work in fashioning distinctive cultural responses to the social experiences of being black in America in the late nineteenth and early twentieth centuries, there seems to be no special evidence for the view that he grew up and formed his outlooks and approaches in an environment characterized above all by racially determined cultural demarcation. Nor is there any evidence that his sound world was controlled and dictated by jazz. This is not, of course, to say that being black was of no significance in the story of his life and his art or to question the obviously profound influence of jazz on his life. But his sound world seems to have been formed by exposure to the products of a wide range of musical cultures (embracing nineteenth-century parlor songs and popular classics, marches, ragtime, and Protestant church music), and he seems to have been equally interested in most of them, regardless of their provenance.

It is no less true of Tatum than of his contemporaries in music that his musical environment was not restricted to one style of music (it may have been particularly significant for him, as I shall consider later). It is doubtful whether sealing oneself off musically, severely restricting one's musical experience according to taste, was ever a realistic option in the United States, but by the 1920s, with the arrival of commercial radio broadcasting, it was almost impossible. Tatum grew up during years of profound change in the ways in which music could be experienced. Individuals--both musicians and nonmusicians--constructed their use of music (as they still may do) according to factors in their individual circumstances and to their own particular position in a network crossing many spheres of life. How they did so was affected in considerable measure by factors (operating singly or, more probably, in combinations) such as class, race, gender, religion, mobility, geographical location, and purchasing power. It was also influenced by the activities of agents of provision and by the variety of opportunities and constraints that presented themselves. At the time of Tatum's birth, these agents of provision included not only songwriters, performers, and instrument makers but numerous intermediaries: publishers, retailers, theatrical entrepreneurs, and venue owners. It also included a young but steadily growing body of record companies.

In Tatum's early years, the world of commercial popular music, although increasingly complex, continued to be dominated by a conjuncture that had been in force for some time. That conjuncture placed class and economic factors in coalition with the activities of a particular grouping of intermediaries--piano manufacturers, sheet-music publishers, retailers, performers, and theater owners--while leaving space for other factors, such as race, as partial variables. It may well have been that one factor in Tatum's personal makeup, his visual impairment--as a result of which printed notation was of limited personal value (although he did learn to read music)--in effect gave him a route out from at least part of this particular dominant alliance and encouraged both his ear and his practice of experimentation. In any case, by the time he was a teenager the old coalition had substantially broken apart, to be replaced by one in which the new media of recordings and radio played a dominant role. In this new situation, the power of visual technology in alliance with sound had been challenged by that of sound alone; and by the time visual technology reasserted itself in the movies, sound alone was too well established to be threatened by a competing technology.

In the new environment, the factor of race was both more and less important. It was more important because the intermediary role played by sound-based media was far more suitable to vernacular African-American musical practice than was that of the print and notation-based media, which had only grafted African-American music onto their production in a small way--and, it is often argued, at the price of considerable distortion. Now, African-American music was not only a significant element in the industrial production of music (i.e., in the record companies' catalogs), but much of it was marketed in such a way (i.e., primarily at a black audience) as to emphasize that it was part of black culture. But race was also less important because, while one product of the sound media--records--was produced and sold within a structure that accepted race as a distinguishing factor, the other--radio--relied on a technology that offered the opportunity for those who controlled it to substantially ignore boundaries such as those determined by class, race, and geography. Radio airwaves could penetrate the lives of anyone who met the one basic requirement--access to a set--and ignore other previously decisive differences. A key part of its method was based on the recognition that, once a home possessed a radio, access to music required no further investment. Radio used this to extend its audience, not by sustaining the heterogeneous appeal favored by the record industry (although radio programming catered to some degree to what later became known as niche markets) but by making fewer products work harder.

There is no doubt that in Toledo Tatum heard African-American music on the radio in the shape of performances by Duke Ellington and Fats Waller; and, according to his biography, he himself broadcast on radio there. Equally, however, there seems little doubt that radio provided Tatum, like many others, with an unprecedented potential exposure to the stream of popular music that dominated the airwaves. To give but one example, he is known to have been a devotee of the pianist Lee Sims, who made many radio appearances in the late 1920s, featuring his particular orchestral approach to popular piano music and his particular fondness for runs and arpeggios (Lester 1994, 123-126; Spencer 1978).

Tatum was not the only black musician of the time to demonstrate a great familiarity with the dominant popular music tradition of his day, but I suggest that in few if any other contemporary black musicians within the compass of jazz was that tradition both so essential to his own self-expression and so inescapably present in its execution. No doubt, on a day-to-day level he gained that familiarity by a number of routes, including recordings of other black musicians performing their versions of the songs of the day. But perhaps as a consequence of his partial blindness, he was an avid radio listener (he was particularly fond of sports broadcasting). He also had remarkable musical memory. It is surely more than mere coincidence that the particular popular tunes on which he constructed his music (or perhaps more accurately, to which his music was a response) were almost all staple fare for radio orchestras. A partial sample of tunes performed by Tatum in his career shows that a significant proportion date from the 1920s and early 1930s, before he left Toledo for New York (see Table 1). If Tatum's recordings are anything to go by (and given his personal reticence, little else exists from the man himself), these tunes played a major part in the soundscape of his life.

Hegemonies, Dialogues, and Intertexts

We will return to the subject of Tatum's sound world later. Let us now turn once again to his reputation. I want in particular to identify some of the specific features that critics have noted as characteristic of his method and to consider whether--and if so, how and with what consequences--we can reassess these in light of Tomlinson's enjoinder to remember the "complex dialogues of self and other in which culture is created."

One of the most common methods of summarizing how Tatum differed from other jazz musicians has been to place him in relation to other jazz pianists, to identify strands of influence, and on that basis to establish his uniqueness. British critic Ray Spencer (1966, 11), for example, provides a particularly concise instance of this approach: "To dissect Tatum's style reveals the foundation of James E Johnson, Fats Waller and Earl Hines. To this Tatum added exciting new dimensions in chord structure on which he superimposed those sparkling runs and arpeggios.... The most common Johnson-Waller trademark found in Tatum is the stride left hand and the use of thirds and broken chords in the right hand. The Hines influence revealed itself in passages of hard-hitting tremolo octaves." Spencer goes on to identify and speculate about another source, from outside of the sphere of jazz: "The runs and arpeggios Tatum introduced into his solos originated from his study of the classics. He cleverly adapted them to the jazz idiom and made them an integral part of his style. I think it is likely that Tatum's knowledge of advanced harmony came from the same source" (11). Spencer's choice of "exciting" to describe Tatum's harmonic approach and "sparkling" to describe his introduction of runs and arpeggios suggests that he has no particular problem with either of these features of Tatum's style. He provides a number of examples of harmonic innovations, including how Tatum "used chromatically a series of chord changes of minor sevenths to dominant sevenths in his 1934 recording of `When a Woman Loves a Man'" (11). Other critics and musicians have shared Spencer's enthusiasm (for example, Teddy Wilson, quoted earlier). But by no means is everyone equally content with the particular matter of the "sparkling runs and arpeggios." Gunther Schuller (1989, 480), for example, speaks of Tatum's "careening arabesques." Schuller, in fact, is a skeptic in matters regarding both harmony and "ornamentation," declaring that "Tatum's `originality' was undercut by the redundancy with which he used certain harmonic and ornamental devices" (481). As a result, in his view, Tatum's art is more accurately designated a craft, for it "remains eclectic, largely predictable, and surface" (478).

At stake here for Schuller is the inviolability of a certain notion of artistic value, as applied to jazz. Tatum's approach, he states, means that he "was not truly speaking an improviser" (481). Whereas outstanding jazz musicians create original material out of old, Tatum "simply appropriated that which already existed and elaborated upon it" (480). In sum, "His was a profuse art, so abundant that its problems were not ... how to attain greater technical control of his materials, but, rather, how to channel his superior gifts into a more deeply expressive and creatively original language--challenges which, if his career is seen in the long view, he did not entirely meet" (477). Whether Schuller was consciously echoing the earlier critique offered in 1959 by the French critic Andre Hodeir, there are striking similarities. Hodeir, too, complained about (among other things) ornamentation, finding that "overall, too many decorative effects stand in the way of the continuity and even the unity of a musical discourse" (Hodeir 1962, 179). While the particular target here is the same, the ideal against which Hodeir believes it should be measured is even grander than Schuller's. In writing of Tatum's "lack of ambition" in remaining close to the melody, Hodeir contrives a total put-down, the more severe for the faint praise that precedes it: "Equipped with greater technical means and a better imagination than one finds in any other pianist, he has an easy time doing superbly what I would have preferred not to have seen him bother with at all" (176).

Tatum's fondness for quotation has also periodically irritated critics. British writer Michael Gibson (1960, 3), for example, deplored the "long series of inapposite, showy and altogether incomprehensible ... quotations" that Tatum began to include in the mid-1940s. "Why," he asks impatiently, "should My Old Kentucky Home ... be deliberately added as a rehearsed coda to Someone to Watch Over Me?"

From these critiques, we can identify a set of practices that are judged to be fundamentally problematic: elaboration, ornamentation, decoration, appropriation, eclecticism, quotation, and preservation of the melody. Frequently linked to these are a number of adjectives: surface, profuse, and abundant. Absent (in the critics' terms) are deep expression, creatively original language, and unity of discourse. It is clear that the critics are judging these practices by their lack of fit to an already established idea, in effect, to what a musical ideal (one notable here by its absence) should contain. My concern here is partly with what this says about the ideal but chiefly with the evident lack of interest in the possibility of alternative perspectives and interpretations.

Although most jazz criticism covering the period of Tatum's career, including the school of writing represented by these examples, acknowledges the existence of a relationship between the original piece and the jazz performance, it shows little or no enduring interest in that relationship. Indeed, apart from noting the harmonic changes that the original provides, the role of the prior musical data is regularly reduced to a minimum. The aim of identifying how closely the musician in question fits the ideal is not one that allows much space for dealing with relationships of this kind. In Tatum's case, it is striking that the factors identified as problematic all place the existence of a relationship in a more prominent position: elaboration, decoration, and ornamentation all concede a considerable level of importance and persistence to the prior music on which these practices are worked; eclecticism, appropriation, and quotation point to interactive processes between two pieces of music; and the preservation of melody suggests a relationship based on coexistence, in which one piece is embedded in another.

If the kind of jazz writing in which Art Tatum is typically featured has little interest in such relationships (whether it is broadly pro-Tatum or not), the same is clearly not the case with theoretical ideas developed in other critical quarters, most notably in the poststructuralist perspectives found most obviously (but not uniquely) in semiotics and in literary and cultural studies and collected together under the term intertextuality. As Richard Middleton (2000, 61) has noted, the key feature of this "rather modish term," namely, that "all texts make sense only through their relationships, explicit or implicit, with other texts," means that it embraces a range of techniques, including borrowing, parody, quotation, allusion, glossing, and punning. The basic requirement that unites them, that a text refer to other texts, results in a concept that "pushes against the tendential self-sufficiency of `works'."

Finding the one term intertextuality inadequate to cover the variations it reveals in practice, some theorists have attempted to develop subcategories. One such is Canadian musicologist Gerard Genette. Genette's ideas, set out in his 1982 study Palimpsestes--in itself a suggestive title for our purposes--have been very usefully laid out and applied by fellow Canadian Serge Lacasse (2000, 35ff), in the field of popular music. As Lacasse explains, in Genette's formulation transtextuality becomes the umbrella term. For Genette, transtextuality means (to quote from Lacasse's summary) "any type of relation, explicit or not, that may link a text with others" (36). That, basically, is other people's "intertextuality." Beneath the umbrella term are five subcategories, identifying five different types of textual relationship: intertextuality, hypertextuality, metatextuality, architextuality, and paratextuality. The first two of these concern us here. Genette uses intertextuality to mean "the effective presence of a text within another" (38), that is, the "copresence" of two texts. Hypertextuality refers to "any relationship uniting a text B (the `hypertext') to an earlier text A (the `hypotext'), upon which it is grafted in a manner that is not that of commentary." (A commentarial relationship is the main characteristic of metatextuality.) "In other words," states Lacasse, "a hypertext is a result of some kind of transformation or imitation of a hypotext"; it involves practices that "aim at producing a new text out of a previous one." Intertextuality, by comparison, involves practices that "aim at including elements of a previous text within the present text" (37; emphasis added).

It is not difficult to relate both of these terms to a great deal of jazz. In approaches that are based on taking a preexistent musical idea--be it a tune, a phrase, a scale, or a rhythm--and creating a performance that derives elements from that idea, there is inevitably a "copresence" in some form, even if the degree of departure from the original varies and even if, in some cases, what is being engaged with is an "original" that has already undergone a degree of transformation in the musician's mind. "The effective presence of one text within another" (intertextuality) is a particularly concise way of summarizing the ongoing presence of the original musical idea, whether that presence is notable as harmony, melody, rhythm, or all three. It also places the phenomena of quotation and allusion (often a source of irritation for critics, whichever musician is responsible) in a wider context. Similarly, it makes a great deal of sense to describe many jazz performances as being directed toward "producing a new text out of a previous one." In the jazz performances of some of Tatum's contemporaries, who like him learned their trade in the 1920s and 1930s and also based much of their music on popular songs of the day, it is clearly appropriate to speak of the creation of a new text from an old one. One thinks of Coleman Hawkins' celebrated 1939 recording of "Body and Soul" or of numerous recordings by Lester Young in which he seems able to breathe new life into the dry bones of a song, such as his 1945 recording with Nat King Cole of "Mean to Me." (Young's approach, like Tatum's, is also notable for its fondess for playfulness, a characteristic Genette and Lacasse place under the heading of hypertextuality.)

In much of the jazz performed in Tatum's adult life, intertextuality and hypertextuality are both present, but the combination is rarely an even one. To put it another way, not all jazz performances that are predicated upon the presence of one text within another appear to want to transform that text in order to create a new one out of the encounter. Equally, not all jazz performances that arise out of a separate text have much interest in that text beyond the raw data it provides (which usually means chord progressions).

The distinction is not necessarily one that can be drawn by putting individual musicians on one side of the line or the other: they are quite likely to use both approaches on different occasions. Instead, we can speak of intertextual and hypertextual tendencies. Many of Sidney Bechet's recorded performances, for example, can be described as predominantly intertextual: the original melody dominates a recording such as "Sweet Lorraine" (1940), whose beauty lies in a combination of melodic variation, subtleties of rhythm, and above all, the sonorities obtained by the duetting frontline musicians, Bechet and trumpeter Muggsy Spanier. Other Bechet performances, such as the celebrated "Blue Horizon" (1945), are more transformative: Bechet's miraculous variations of sound and musical line produce a recording that has been much admired as, in effect, a new text.

One can identify intertextuality and hypertextuality in Art Tatum's music also. Intertextuality in Genette's sense is present in his fondness for quotation that so annoyed Gibson. (A good example is "Nobody Knows the Trouble I've Seen," interpolated regularly into "Body and Soul.") He also uses an approach that is closer to allusion, hinting at another melody. (1) But more important, the term copresence draws attention to the frequent prominence of the original in Tatum's approach. Hypertextuality for its part seems clearly present in the way in which he takes an existing popular song and effects some kind of transformation. This is apparent in his custom of altering the harmony of a song during his performance (Benny Green once described a Tatum treatment of Ellington's "Sittin' and a-Rockin'" as "skittering with grace down unsuspected harmonic corridors" [Green 1975]). But the transformation he effects is often much more extensive. For example, in his 1953 Los Angeles performance of "Love for Sale," the original remains recognizable melodically but in no other way, finding itself involved in a complex sound environment in which harmony, rhythm, tone color, and mood shift rapidly, humorously, and sometimes dramatically. As Lester (1994, 130-131) has suggested, such transformations also have an architectural quality, where "large thought-out patterns (give) coherence to details that might seem fragmentary."

Despite these transformations, it may not be entirely appropriate to speak--or speak only--of the creation of a new text. Tatum can be listened to and thought about in this way, but I suggest that certain factors make it uncomfortable as an overall interpretive stance. Here, we return to the "problematic" features of Tatum's style. I want to focus on two in particular. The first is the persistence of melody. Although there is much variety in how Tatum goes about his work, certain features are common to many recorded performances. One is the way in which the original melody retains a distinctive presence in the midst of much, often quite frenetic, activity. It does not, however, dominate. It stimulates activity, and often appears to be the center of activity, without dictating it. Although often fragmented, it remains as a kind of hard core, involved in the activity but relatively unaffected by it. No matter that Tatum subjects it sometimes to a harmonic rewrite, no matter how much he sometimes varies it--so that, employing the everday language often used to describe musical interpretation, one could perhaps say he "gets inside the music"--he appears above all to want to keep the original melody out there where he can hear it and talk with it. Occasionally, he seems to want to rid himself of it but cannot. In these circumstances, it makes little sense to speak of Tatum absorbing the original, or interpreting it, or re-creating it--or indeed overpowering it. It is, in the end, an immutable companion.

The second "problematic" feature embraces the three related terms elaboration, ornamentation, and decoration (for convenience, I subsume them under the term ornamentation). In broad terms, much musical ornamentation acknowledges the dominant presence of the idea being ornamented and does nothing to challenge it. Indeed, in baroque music it was commonly thought that ornamentation of a melody enhanced the melody's character. Other types of ornamentation are designed to add an element of display. This display does not attempt to affect the musical idea or indeed any subsequent development, extension, or other teleological occurrence in which the idea is involved; instead, it draws attention to itself--sometimes modestly, sometimes with no modesty. In Tatum there is ornamentation of the melody in the "traditional" sense, by means of devices such as appoggiatura; equally, there are instances in which the ornamentation may seem to be engaged in self-display. But--and perhaps more important than either of these--there are also the devices that are associated with ornamentation but to which Tatum gives a space, and role, of their own. Especially prominent are the runs--"the frequent left-hand, right-hand, and two-handed runs, all superlative and incomparable," as Samuel Floyd (1995, 112) has described them.

Although we could discuss these features separately in this way, it does not entirely make sense to do so. For the focus is not on the persistence of melody pure and simple but on the way that persistence is handled; and one major way in which Tatum handles it is, almost literally, by the ornamentation. Examples are plentiful. In both the 1945 V-disc recording of "Lover" and the 1955 recording of "When Your Lover Has Gone," Tatum uses the ornamental devices to walk around the melody. They keep a bit of distance from it, then come close to it; they talk with it, toy with it, spin off it, push it aside, and then let it back in; perhaps we could say, they find ways to live with it. It is a rich and restless dialogue.

The dialogue between ornamentation and melody is not the entire story, however; ornamentation does not work with (at, alongside) melody alone, any more than does any element in Tatum's style in its relationship with anything else.

Another key element is his famous left-hand bass patterns. Tatum's left hand has been much admired, even by some of his sternest critics, but his decoration has not; yet both are part of the same enterprise.

Beyond the Text I: Space and Theater

How can that enterprise be described? In a formalist sense it is possible, and important, to build a picture of the relationship between structure and detail in a Tatum performance and so obtain the evidence that will allow us to confirm Howlett's (1983, 236) unexceptional but significant comment that "there is often a variety of musical activity taking place simultaneously." The excursion into intertextuality and hypertextuality is not only highly suggestive for styles of jazz that were current in Tatum's lifetime; it also enables us to begin to see beyond formalism into the complexity of dialogical relationships in Tatum's music, relationships both between musical texts (his and the preexisting ones he chooses) and within the ones he creates. But then what?

I confess to feeling uneasy about taking theories developed for the study of literary texts and applying them too rigidly to music, especially if the process results in excessive attention being paid to certain aspects and too little to others. Theories based on notions of textuality--however broadly the notion of a text may be understood--almost inevitably focus attention on the results of activity and away from the activity itself. One possible solution may be found in the work of Ingrid Monson (1996), who has proposed the term intermusicality as a musical equivalent of intertextuality to describe the ways in jazz that interplay is conveyed aurally through musical sound itself. More than can be expected from intertextuality, intermusicality draws attention to interactive activity in performance, whether between musicians or within the music. (2) Monson uses intermusicality in the context of group interaction, but can we adapt the term for Tatum?

Such highly accomplished pianists as Oscar Peterson and Hank Jones have jokingly referred to their impression that Tatum must have more than two hands or that there must be more than one person playing (Lester 1994, 44-45; Pullman 1996). One can approach the often-bewildering intersecting activity that Tatum presents as "a variety of musical activity taking place simultaneously"--in other words, as a complex layered treatment of time. But we can also hear the various lines (whether such a line is an ornament, a rhythmic figure, an ostinato pattern) as creating patterns of movement across space. In this space, as prowling left-hand figures, dramatic runs, fragments of melody, and heavy chords interact and intersect, they create the impression of multiple presences, each with a role to play.

The theatrical metaphor is not accidental. It is tempting to think of any jazz soloist as engrossed in an abstract challenge with the many parameters of musical expression--perhaps especially tempting when the soloist is blind or partially blind--and clearly Tatum set himself enormous technical challenges with his material. But a Tatum performance is not only about abstract challenges. The way in which he preserves the persona of the tune and introduces all manner of gesture and movement around it, and conversation with it, suggests a conception that is dramatic, even theatrical.

Space and movement, gesture and dialogue between characters are all readily apparent in a recording such as "Please Be Kind" (1955). Here, Tatum employs a favorite tactic of a slow introduction, not out of tempo but with no stated rhythm. Right- and left-hand phrases toy with the melody in an exploratory fashion, as if in an initial process of getting acquainted. Then the left hand introduces a regular rhythmic movement, and it is not too much to say that the multiple presences begin to dance. A choreographed theater of the mind it may be, but one with very realistic touches. The left-hand figure is a bit discordant, somewhat disgruntled, out of sorts, or uncooperative; then it gradually comes around.

Thinking of his music in this way enables us to begin to put into context the widespread unease that is often felt about ornamentation in Tatum's music. Unease with ornamentation and decoration--with what Schuller had in mind when he termed Tatum's art "profuse"--has a long history, especially in Protestant cultures. It links to fear of display and progresses readily to accusations of insincerity. Historically, part of this fear has often been directed at theatricality. Jonas Barish (1981, 117) sees the hostility characteristic of antitheatricality as based in fear of the negative effect on order and stability of an emphasis "that prizes ... exploration, flexibility, variety and versatility." As such, antitheatricality embodies "an ideal of rectitude," central to which is truth to oneself (echoing the Puritan-derived fear of dissembling). This ideal of rectitude is pitted against an "ideal of plenitude." In the case of jazz writing, the two ideals are not so diametrically opposed; the components of plenitude (exploration, flexibility, variety, and versatility) are seen as central to the music. But rather than leave them in tension, the workings of the jazz canon have reconciled them in a particular way. Plenitude may be reconciled with rectitude if it is grounded in practice that has internal organic unity and if it is the authentic reflection of the inner person (whatever his or her character). This approach has great difficulty when faced by Tatumesque ornamentation, which seems less preoccupied with organic unity or inner self. On top of that, the multiple presences that populate the Tatumesque world are not shy. They have no fear of display; it is part of their way of life, part of what enables them to relate to one another.

In the context of an Art Tatum performance, the concepts subsumed under Genette's transtextuality and Monson's intermusicality lack the ability to suggest another set of relationships: that between the Tatum performance itself and the performative activity it represents. In other words, we are in the presence of the "word facing both ways." Mikhail Bakhtin's phrase acknowledges both the dialogical relationships between the original musical ideas (especially melodies) and Tatum's performance, and those within the performance, but it also draws attention to the fact that that performance faces outward toward another.

Beyond the Text II: Dualities and Sound Worlds

This exploration so far has, I hope, done something to reveal the dialogical complexity in Art Tatum's music, but we still need to examine that complexity in a broader cultural sense. One can take a cue from Monson (1996, 129) and note that intermusical relationships "may be intercultural relationships as well." She is speaking of the way in which jazz musicians draw on music from outside of African-American music: "The cultural knowledge of African American musicians includes familiarity with both `black' and `white' music, and it is upon all of this knowledge that a musician draws in the act of performance. This is what I mean when I say that music making is an active participant in cultural discourse" (130).

As Monson acknowledges and discusses, the concept of signifying has become noted for its ability to make a link between ideas derived from intertextuality and the specificities of the black experience in response to its complex cultural encounters. The now well-known term requires no detailed description or analysis here, but some key ideas need to be emphasized. In the seminal text by Henry Louis Gates (1988), the folk tales of the trickster figure, the signifying monkey, and their African predecessors are used as a key to develop an argument that there is a centrality in African-American cultural practice of a set of rhetorical interpretive strategies that are based around figurative troping devices and that proceed not by passing on information but, like the monkey, by manipulating it. Thus, "One does not signify something; rather, one signifies in some way" (54).

A key element in signifying that others have picked out from Gates's formulation has been the act of working transformatively on preexisting material. This is seen clearly in Floyd's (1995, 8) summary of how the concept translates into music: "Musical signifyin(g) is the rhetorical use of pre-existing material as a means of demonstrating respect for or poking fun at a musical style, process, or practice through parody, pastiche, implication, indirection, humor, tone play or word play, the illusion of speech or narration, or other troping mechanisms."

Use of the term developed originally within African-American vernacular culture, where it meant, as anthropologist Roger Abrahams (1970, 264) neatly summarized it, "a language of implication." From this, Gates deduced that the vernacular culture had "decolonized" the term from its original English usage. In the process, as Gates (1988, 50) put it, the semantic register had been replaced with the rhetorical, the syntagmatic axis with the paradigmatic. "Everything that must be excluded for meaning to remain coherent and linear comes to bear in the process of Signifyin(g)." (3) But at the same time, in Gates's theorization, the shadow of the "other" earlier meaning of "to signify" remains: the fact that the act of signifying, in whatever context, could be taken as part of a syntagmatic chain of meaning is part of the ambiguity.

Many scholars have noted the duality that this involves, as well as its possible indebtedness to W.E.B. Du Bois' famous double consciousness metaphor. While acknowledging this, Gates also draws Bakhtin into the frame. Bakhtin's concept of the double-voiced utterance allows the important perception that an utterance can have a new semantic orientation inserted into it and still retain its earlier orientation; and that utterances can be inscribed with each other's meanings (always allowing for the unequal power relations often at work in each context). This in turn enables Gates to encapsulate signifying as "black double-voicedness" (Gates 1988, 50-51).

The concept can appear overarching, but two related factors give support to the idea that signifying may be an appropriate term to use in exploring African-American musical culture. First, although the term is no more peculiar to music than intertextuality and related terms, it distinguishes itself in being a vernacular term that has been adopted and adapted for use in academic discourse. For Gates, this is not accidental; he insisted that there was no academically imposed distance between his theoretical approach and vernacular usage and that he was identifying "a theory of criticism that is inscribed within the black vernacular tradition" (xix). And indeed, the word does occur in vernacular speech in musical contexts (for example, in the title of a Count Basie recording).

Second, Gates submits that, although parallel types of interpretive strategy exist in other cultures, a distinctive term is needed because African-American experience is different. The idea of signifying keeps us alert to the fact that, in the context of North America, faced with the persistence and insistence of the dominant culture, the specific cultural practice of manipulating cultural data in ways that mixed semantic and rhetorical registers--a practice that African-American culture no doubt shared with other cultures--acquired a particular significance. Across the years, it came to be used with increasing sophistication not just as a survival mechanism but as a creative response.

In many respects, the term signifying is an appropriate one to use in discussing Art Tatum, and it is the context in which he is briefly discussed by Floyd (1995, 112-113). Although signifying lacks the subtle variations of the nest of terms brought together by Genette under transtextuality, it comes close to Monson's intermusicality in its stress on highlighting activity in preference to the consequences of activity; indeed, as a verb rather than an abstract noun, it embodies that preference within itself. In discussing the applicability of signifying, Monson (1996, 87) cautions against extending its use to the works of individuals. What is lost, she maintains, "is a sense of how signifying as an aesthetic developed from interactive, participatory, turn-taking games and genres that are multiply authored." But far from being a reason not to extend the idea to an individual such as Art Tatum, that very description seems to encapsulate much of his music. Single-authored though it may be, his music could be said to be built around re-presenting "interactive, participatory, turn-taking games" in which the multiple presences do their own dialogical authoring.

But there is more. Signifying adds to the frame the crucial element of cultural particularity, connecting dialogues of self and other to the specific experiences of being an African-American artist. Tatum's music as we have discussed it not only explores dialogic relationships between texts, between text and performance, and between performance and performance, but it also explores relationships with the dominant culture. In basing his music almost entirely on that culture's Tin Pan Alley and Broadway repertoire, Tatum could be said to be signifying on some important representatives of that culture's badge of identity. In this context, the fascinating encounters between the melody, almost stubborn in its persistence, and the apparently endlessly varied movement and gesture around it constitute eloquent statements in the tradition of "black double-voicedness." The mixing of semantic and rhetorical registers is readily apparent in a performance such as "Danny Boy," recorded live in 1955 in the home of Hollywood musical director Ray Heindorf. Here, Tatum plays with and gently parodies the famous Irish tune, while at the same time loving it to death. One suspects, listening to the reactions of his audience, that they got the idea too.

Yet once again, we should be wary of drawing the interpretive string too tightly around the object of study. Monson (1996, 125) reminds us that "[e]ach individual has a personal listening world that intersects to a greater or lesser degree with those of other participants in a particular musical tradition, but no two people are likely to have exactly the same sound worlds." These diverse sound worlds are inevitably structured: there are always forces at work that tend toward unification, centralization, and standardization, and there are forces that encourage multiplicity and, in Bakhtin's term, heteroglossia. But there are no absolutes. Certain patterns, certain power relations between sonic forces, may be more likely in any given context (such as African-American culture) and may have a powerful role to play, but ultimately, the way particular forms of musical expression are structured in relation to one another also varies from person to person. Monson quotes approvingly from the poet Elizabeth Alexander, who, she says, "has proposed the term collage to describe the multiplicity of voices that coexist in African-American identity. Alexander ... explicitly moves away from a dichotomous opposition of black and white to a perspective that includes multiple voices. She rejects the notion that African Americans speak from a state of `spiritual and cultural schizophrenia and self-division' in favor of a conception that `maps a theoretical space in which the myriad particulars of identity can reside'" (100). That is the kind of thing I have been attempting here.

Art Tatum performed most frequently as a soloist and so has often been judged as such, complete with all the trappings of a monological individualist, insufficiently interested in understanding others. But I have tried to suggest that his personal world was replete with a multiplicity of others. The myriad particulars of his identity can only begin to be understood when we accept the heterogeneity of his musical and cultural world, the multiple presences with which he peopled the world he created, and the complex dialogical interplay it drew on and expressed.

Table 1. Publication dates of a selection of tunes featured in Art
Tatum recordings

After You've Gone                    1918
Aunt Hagar's Blues                   1920
(Back Home in) Indiana               1917
Begin the Beguine                    1935
Body and Soul                        1930
I Can't Give You Anything but Love   1928
I Cover the Waterfront               1933
It's Only a Paper Moon               1933
Jitterbug Waltz                      1942
Love for Sale                        1930
Making Whoopee                       1928
Mean to Me                           1929
Mighty Like a Rose                   1901
My Heart Stood Still                 1927
Over the Rainbow                     1939
September Song                       1938
She's Funny That Way                 1928
Someone to Watch Over Me             1926
Stay as Sweet as You Are             1934
Sweet Lorraine                       1928
Tea for Two                          1924
There Will Never Be Another You      1942
Too Marvelous for Words              1937
The Very Thought of You              1934
When Your Lover Has Gone             1931
Willow Weep for Me                   1932
Without a Song                       1929

(1.) In Serge Lacasse's application of Genette's theory to popular music, he provides a number of types of intertextuality and hypertextuality, among them quotation and allusion, parody and pastiche.

(2.) Monson does not attempt to introduce any subcategories along the lines of Genette's sophisticated system, but there seems to be no particular reason why something similar could not be attempted.

(3.) Gates' reason for adopting the spelling "Signifyin(g)" is that it enables him to represent orthographically the fact that African-American culture signifies on the processes of signifcation typical of mainstream culture. As he puts it, "the absent g is a figure for the Signifyin(g) black difference" (Gates 1988, 46).

DISCOGRAPHY 

Bechet, Sidney. Blue horizon. Sidney Bechet: Jazz classics. Vol. 1. Blue Note CDP7893842. Compact disc.

--. Sweet Lorraine. Bechet-Spanier big four. Joker SM3090.

Hawkins, Coleman. Body and soul. Smithsonian collection of classic jazz. Columbia Special Products P6 11891.

Tatum, Art. Danny boy. Art Tatum: 20th century piano genius. Verve 531 763-2. Compact disc.

--. Love for sale. The Tatum solo masterpieces. Vol. 1. Pablo 2310-723.

--. Lover. Art Tatum: The V-discs. Black Lion Records BLP 30203.

--. Please be kind. The Tatum Solo masterpieces. Vol. 10. Pablo 2310-862.

--. When your lover has gone. The Tatum solo masterpieces. Vol. 10. Pablo 2310-862.

Young, Lester. Mean to me. Lester swings. Verve 2610 039.

REFERENCES

Abrahams, Roger D. 1970. Deep down in the jungle: Negro narrative folklore from the streets of Philadelphia. 2nd ed. Chicago: Aldine.

Barish, Jonas. 1981. The antitheatrical prejudice. Berkeley: University of California Press.

Floyd, Samuel A., Jr. 1995. The power of black music: Interpreting its history from Africa to the United States. New York: Oxford University Press.

Gates, Henry Louis, Jr. 1988. The signifying monkey: A theory of African-American literary criticism. New York: Oxford University Press.

Genette, Gerard. 1982. Palimpsestes: La litterature au second degre. Paris: Seuil.

Gibson, Michael. 1960. The paradox of Art Tatum. Jazz Journal 12, no. 10 (October): 3-4.

Giddins, Gary. 1998. Art Tatum (sui generis). In Visions of jazz: The first century, 439-444. New York: Oxford University Press.

Green, Benny. 1975. Liner notes, The Tatum solo masterpieces. Vol. 1. Pablo 2310-723.

Hodeir, Andre. 1962. The genius of Art Tatum. In The art of jazz: Essays on the nature and development of jazz, edited by Martin Williams, 173-179. London: Jazz Book Club.

Howlett, Felicity, 1983. An introduction to Art Tatum's performance approaches: Composition, improvisation, and melodic variation. Ph.D. diss., Cornell University.

Howlett, Felicity, and J. Bradford Robinson. 1988. Art Tatum. In The New Grove dictionary of jazz, edited by Barry Kernfeld, 2:519-520. London: Macmillan.

Johnson, Bruce. 2002. Jazz practices. In The Cambridge companion to jazz, edited by David Horn and Mervyn Cooke. Cambridge: Cambridge University Press.

Jones, Max. 1956. Tribute to Tatum. Melody Maker (November 10): 8.

Lacasse, Serge. 2000. Intertextuality and hypertextuality in recorded popular music. In The musical work: Reality or invention? edited by Michael Talbot, 35-58. Liverpool: Liverpool University Press.

Lester, James. 1994. Too marvelous for words: The life and genius of Art Tatum. New York: Oxford University Press.

Middleton, Richard. 2000. Work-in(g)-practice: Configurations of the popular music inter-text. In The musical work: Reality or invention? edited by Michael Talbot, 59-87. Liverpool: Liverpool University Press.

Monson, Ingrid. 1996. Saying something: Jazz improvisation and interaction. Chicago: University of Chicago Press.

Pullman, Peter, ed. 1996. Liner notes, Art Tatum, 20th century piano genius. Verve 531 763-2 (CD).

Richardson, Joanna. 1976. Victor Hugo. London: Weidenfeld and Nicolson.

Schuller, Gunther. 1989. The swing era: The development of jazz, 1930-1945. New York: Oxford University Press.

Spencer, Ray. 1966. Art Tatum: An appreciation. Jazz Journal 19, no. 8 (August): 6-16.

--. 1978. Liner notes, Art Tatum: The V-discs. Black Lion Records BLP 30203.

Tomlinson, Gary. 1991. Cultural dialogics and jazz: A white historian signifies. Black Music Research Journal 11, no. 2: 229-264.

Wiedemann, Erik. 1955. The Art Tatum myth. Jazz Monthly 1, no. 7 (September): 27-28

Young, Gavin. 1963. Three pianists discuss Art Tatum (and other matters). Jazz Journal 15, no. 11 (November): 22-24.

DAVID HORN is director of the Institute of Popular Music at the University of Liverpool. He is joint managing editor of The Encyclopedia of Popular Music of the World and co-editor of The Cambridge Companion to Jazz.