- The Blues Formula:A Response to Taft
Michael Taft ends his response to my critique (2007) of his book, The Blues Lyric Formula (2006), on a heroic note, describing how, at the 1972 annual meeting of the American Folklore Society, he survived Albert Friedman's criticism of his position on formulaic composition in the blues (2009:80). I personally witnessed several of Friedman's harsh remarks to hapless graduate students at academic meetings and saw how they left permanent scars on some, reduced others to tears, and drove still others from the field of folklore studies altogether. Therefore, I congratulate Taft for his ability to hold his ground and go on to a distinguished career as a folklorist. The beginning of his response, however, is far less heroic, as he tries to make a virtue out of his faulty transcriptions of the lyrics of early blues recordings. Taft admits to a shocking error rate of "perhaps 10 percent of the words in any given song" but claims that his inaccuracies "tended to make my corpus seem less formulaic than it actually was" (2009:75). He states that "transcribing for the purpose of formulaic analysis" is likely to result in an inflated number of formulas, which is why he "tried not to prejudge the lyrics and erred on the side of some mishearings" (75). I feel almost embarrassed to point out that the goal of lyric transcription from recordings is accuracy. After this has been achieved can come formulaic analysis. But regardless of this weakness in his corpus, Taft and I are not in disagreement over whether "the blues is highly formulaic" (76). We have both long maintained this view. Instead, our dispute is over the nature and definition of the formula and how formulas are used by blues singers. I doubt that either Taft or I will change the other's mind, but let me briefly summarize our differences and reply to some issues he raises in his response.
Taft and I both claim that our work is inspired by the definition of the lyric formula first offered by Milman Parry ( 1971:11). While I uphold a lexical and syntactic interpretation, focusing on the first part of Parry's definition ("a group of words which is regularly employed under the same metrical conditions"), Taft upholds a semantic interpretation, focusing on Parry's concluding phrase ("to express a given essential idea"). I believe, however, that Taft is quibbling over semantics. If Parry had used an alternative word such as "clear," "simple," "specific," or "basic," instead of the word "essential," I think we would have no argument over the interpretation of his definition. Parry was not some sort of closeted proto-Chomskyan semanticist, nor was he being "crypto-psychological" (Rogers 1966; Taft 2009:76). Everything that Parry and his student Albert Lord (1960) wrote about formulas shows that these are [End Page 218] expressions that fix some sort of specific image in the mind. Typically they contain at a minimum a substantive word (noun, proper name, or pronoun) and/or a verb form, with one or more modifying words (adjective, adverb, preposition, article, etc.). Thus, a group of words comprised entirely of conjunctions, articles, prepositions, and modifying words, such as "and slowly into the yellow," would not constitute a formula. That is all Parry and Lord meant by "essential," as can be seen by their detailed discussions of how a formula system works.
In his response to my criticisms, Taft uses blues formula variation ("I woke up this morning" / "I got up this morning" / "I rose this morning") to support his semantic interpretation, claiming that the "essential idea" behind these variants is "emergence from sleep" (2009:76). I agree that these phrases are formulaic variants and that they all describe emergence from sleep, but I would not call "emergence from sleep at the beginning of the day" (2009:77) a formula, as Taft does in The Blues Lyric Formula (2006:55), any more than I would call formulas all the other phrases that this idea is capable of generating (e.g., "I awoke early in the day"). The essential elements of this set of blues formula variants are...