4th ed. Compiled by Robert M.W. Dixon, John Godrich, and Howard Rye. Oxford: Clarendon Press, 1997. [xlix, 1870 p. ISBN 0-19-816239-1. $95.]

Robert M.W. Dixon's and John Godrich's discographical efforts for pre-World War II ("prewar") blues and black sacred music have received such respect

from research peers that no challenger has yet appeared. Oddly, none of the three previous editions of Blues and Gospel Records (1963, 1969, 1982) was reviewed separately in Notes, although all three were discussed in Barry Kernfeld and Howard Rye's "Comprehensive Discographies of Jazz, Blues, and Gospel (Part One)" (Notes 51, no. 2 [December 1994]: 501-47). After the third edition went out of print, interest in prewar blues increased dramatically with the platinum-sales success of Sony/Columbia's 1990 reissue of Robert Johnson's 1936-37 recordings. During the last few years, antiquarians have sold used copies of the previous editions for prices from fifty to one hundred dollars. The publication of this fourth edition is timely, and its improvements over its predecessors render it timeless.

Seventy-eight rpm records, the most common recording medium for prewar blues, are the primary source documents from which secondary studies of this repertory, including notated transcriptions, are derived. While collecting data for their discography, Dixon and Godrich (and later, Rye) sought two important elements: the serial number printed on.the record label for retail purposes, and the matrix number/etched near each disc's run-out groove for identification during recording and manufacturing. If the records were not available, serial numbers were gleaned from old label retail catalogs and dealers' lists, and matrix numbers were taken from surviving session logsheets. The majority of the blues records the compilers needed to consult were found not in libraries but in the possession of collectors who recovered them from oblivion. Eventually, the compilers had comprehensive lists of record releases by label, and of recording sessions by matrix number. This information is rearranged by performing artist in alphabetical order.

The range of recording media included has been broadened beyond the 78 rpm record, almost to the point of bursting the work's scope. The previous editions concentrated on disc recordings produced by commercial labels and the Library of Congress. With the method of culling data described above, Dixon and Godrich succeeded in producing a discography of the 1920 - 43 commercial "race records" era, when blues and sacred music were recorded on 10-inch 78 rpm discs by black musicians for black listeners. The Library of Congress field sessions, recorded on large aluminum discs for study instead of retail, are pertinent because they include artists who had previously recorded little or not at all, and because many performances have been excerpted for issue on commercial 78s, LPs, and compact discs in response to consumer interest in African American folk music. As Kernfeld and Rye wrote in advance of the fourth edition's appearance, it "seeks to list all surviving sound documents" (ibid., 534). Although it is wonderful that the compilers have added cylinders, film soundtracks, and surviving radio transcriptions, perhaps the word "records" in the title should be amended to "recordings."

Likewise, the breadth of the kinds of music has been considerably widened, especially for popular and sacred music. Black popular artists like Josephine Baker, who recorded some blues material for labels outside the "race record" industry, are now included. Others who had been mentioned in previous editions as being "of no blues interest" or, in the case of the Fisk Jubilee Singers, as having "little authentic gospel quality," now have their recordings presented in full. Although some social and cultural factors may have been behind these editorial changes, the presence of these recordings will challenge some musical assumptions. Kernfeld and-Rye include among blues' contributions to jazz "a musical form (the blues progression) [and] a poetic form (aab lyrics)" (p. 530); here they must mean the twelve-measure blues form with which the aab lyric scheme has been most often associated. But early blues-men recording before 1933 used eight- and sixteen-measure forms as well, borrowing some of them from sacred music, and throughout the prewar period other twelve-measure lyric schemes such as aaa and the 4+8 verse-and-refrain chorus were used. In the 1963 edition of Blues and Gospel Records, Dixon and Godrich first asked their classic discographical question, "What is a 'blues' or 'gospel' record?" (p. 3). This new edition may lead some readers to seek defining characteristics in the newly added recordings of black dance, sacred, and "pre-blues" music, rather than in the performances of the paradigmatic twelve-measure aab blues form refined on the eve of World War II.

The front matter presents the basic information every reader ought to absorb before consulting this volume. The introduction states the overall intent and the procedures of compilation, and the acknowledgments and bibliography list the discographical collections and secondary sources. Especially important is the "Race Labels" section, which gives profiles of individual recording firms, including notes on studio recording practices, company mergers, and recording trips. (In 1970, Dixon and Godrich published their classic little Recording the Blues [New York: Stein and Day], in which they consolidated their various label histories into a chronological narrative of the prewar blues recording business.) A digest (provided by John Cowley) of the Library of Congress field recordings, and notes on the radio transcriptions and foreign recordings cited in the performer entries, round out the introduction.

In the discography itself, a good example of the procedures and choices the compilers had to make is the entry for legendary Delta bluesman Son House's 1930 Paramount label recordings (p. 404). For the main access point, the professional name as it appeared on record labels is used, with the real name when known (in this case, Eddie James House Jr.) given immediately after. Full performer and instrumentation details and recording location follow; in this entry, "Grafton, Wis." refers to the Paramount recording studio where House had traveled with Charlie Patton and two additional Mississippi blues musicians. The 28 May 1930 session date assigned by the discographers is too specific and perhaps too early. The Paramount session logsheets were lost after the label folded in 1932, so recording dates have had to be conjectured from surviving retail lists and the "L" matrix number series used for Grafton sessions. The 28 May date is taken from a test pressing for Irene Scruggs's "You've Got What I Want," Paramount matrix L-349-2 (p. 783), sixty matrix numbers before House's first recording. Yet the drought of which House sings in "Dry Spell Blues" (and of which his companion Charlie Patton sang in his own "Dry Well Blues") had not yet begun that May. The date given in the first edition, "c. July 1930," is more reasonable, as by then the drought was in full heat. The recorded performances follow in matrix number order (also chronological, assuming the matrices were numbered in the order they were recorded), with the matrix numbers given on the left side of the titles and the serial label numbers on the right. The first performance listed in this session entry, which lacks the "L" matrix number, is the unique Paramount test pressing (now kept by Yazoo Records) of House singing lyrics about walking to a Blind Lemon Jefferson "moan" melody; where the matrix number should be there is an enigmatic inscription reading "9/2#1." The Yazoo and Document compact-disc releases of this recording are titled "Walkin' Blues," although its melodic form is different from House's 1941 Library of Congress performance tided "Walking Blues" and Robert Johnson's 1936 "Walking Blues," the latter two using House's "My Black Mama" theme instead. Most of the records released before 1943 are cited only by their prewar 78-rpm issues.

The next record listed in the House Paramount entry is "See That My Grave Is Kept Clean," a tune best known from Blind Lemon Jefferson's 1927 hit record that is here given the odd designation "P[ara]m[ount] unissued?" This curious inclusion apparently comes from House's 1965 recollection to Al Wilson (Blues Unlimited Collector's Classic 14 [October 1966]: 5) of having made "Mississippi County Farm Blues" using that Jefferson melody. Although a copy of that studio record has not been found, House's claim of developing such a song may be confirmed by examining his "County Farm Blues" recorded for the Library of Congress in 1942. Yet until a commercial copy of Paramount 13096 is found, the relation of "Mississippi County Farm Blues" to the supposed "See That My Grave Is Kept Clean" will be unverified, as will be the flip side "Clarksdale Moan," which may turn out to be the "Walkin' Blues" test pressing discussed above. (For more on these performances, see my "Blues in the Round," Black Music Research Journal 17 [1997]: 3-36, and my "Son House's 'Clarksdale Moan' Considered," Tri-State Blues 2 [1997]: 16-17.) It would be helpful if the compilers marked by asterisk performances that have not survived in any commercial or test copy, which is true of many records released between 1930 and 1933, the worst years of the Great Depression. The remainder of the records ("My Black Mama," "Preaching the Blues," and "Dry Spell Blues") are presented with full details in matrix-number order. The session entry is completed with a listing of blues artists recording between House's studio takes. In this discographical format, a great amount of information is synthesized, and the flaws tend to be errors of inclusion.

Use of this volume is eased by its size (6 x 9 1/2 2 1/4 in.) and layout. Though somewnat bulky for carrying from shelf to desk, its paper thickness, signatures, and binding enable the pages to lie flat, making it ideal for keeping open during long research projects. On the pages themselves, the type is easy to read; the wide margins allow for annotations, corrections, and checkmarks.

The outstanding feature of the back matter is the title index, which did not appear in the first three editions. This index not only facilitates searches by title, but provides clues to finding multiple recordings of a single blues (although as seen in the "Walkin' Blues" example, the final authority is in the playback of the various recordings). Indexes to vocalists and accompanists have been refined by research appearing since 1982. The second edition's section of "microgroove issues" was dropped because of the flood of LP and compact-disc reissues. (Those wishing for a comprehensive listing of prewar blues and sacred performances on compact disc should write for a catalog from Document Records, Eipeldauerstrasse 23/45/5, A-1220 Vienna, Austria.) Future additions and corrections will be published in the leading British blues magazine, Blues and Rhythm: The Gospel Truth.

Blues and Gospel Records, 1890-1943 retains its eminence as a blues reference and discographical model. One hopes that Brian Rust's Jazz Records, 1897-1942 (London: Storyville, 1969) will be refined and republished to match its high standard of achievement, and that a prewar country (or "hillbilly") discography will be compiled; these would produce many cross-references to Blues and Gospel Records for those sessions when blues, sacred, jazz, country, and ethnic musicians took turns recording. But in itself the work will reinforce its grounding influence. With the recurring misunderstandings of black music, the much needed reappearance of Dixon, Godrich. and Rye's grand opus is an improved - even daring - guide to the recorded sources.

EDWARD KOMARA University of Mississippi