In a book by Dr. Sigmund Spaeth "A History of Popular Music in America" (Random House) 1948, on page 612, Spaeth suggests Gambler's Blues (St. James infirmary Blues) dates to 1899. No composer credit is given indicating a "traditional" tune.i1  

St. James' hospital was founded about the time of the Norman Conquest for "maidens that were leprous."  King Henry VIII took possession of the hospital and it was rebuilt as St. James palace in 1533, becoming the London residence of British sovereigns from 1697-1837.  The song, whose origins probably stem from The Unfortunate Rake , an English folk song from the late 18th century, is a later adaptation of Gambler's Dream from the 1890s.  According to New Orleans historian, Jack Stewart, St. James Infirmary was a temporary facility during the Civil War, but according to Al Rose, author of Storyville, New Orleans, despite a reputed association with St. James Methodist Church, which may have offered first-aid services and modest hospital facilities, the song has no connection with New Orleans whatever. i2

"St. James Infirmary," it turns out, is an offshoot of an extraordinary song cycle that is the subject of a 1960 Folkways Records release called The Unfortunate Rake: A Study In The Evolution of A Ballad, containing 20 songs and extensive notes by Kenneth S. Goldstein. I have, needless to say, purchased this item. Goldstein writes that the oldest published text from the "Rake" cycle was "collected" in 1848 in County Cork, Ireland, "from a singer who had learned it in Dublin in 1790." The song may have been "in tradition" for years prior to that, but it's obviously impossible to say. (He also says St. James Hospital was in London, and treated lepers.)

The disc includes one recording based on lyrics printed on a 19th century broadside. The singer recounts "a-walking down by St. James Hospital" one day and running into a friend, who was "wrapped up in flannel," despite the warm weather. The friend blames his troubles on "a handsome young woman." It seems that he knew this woman rather well, but there was something she didn't tell him, and if only she had, "I might have got the pills and salts of white mercury." This refers to treatment for venereal disease. "Now I'm cut down in the height of my prime," the unfortunate rake explains, proceeding to make requests relating to his funeral ("Get six your soldiers to carry my coffin, six young girls to sing me a song…")

The next several tunes on the disc are variations on this story, with the lyrics rearranged in various ways. One difference is that most are explicit that the young man is a soldier or sailor, and none are any where near so explicit about what exactly his problem is. In fact they're all extremely vague — it's just a young man who is "cut down in his prime" for reasons that' aren't clear. Sometimes, as in "Bad Girl's Lament," the ballad is about the woman, but basically follows the same pattern (an early mention of St. James' Hospital, a closing request for "Six pretty maidens with a bunch of red roses, Six pretty maidens to sing me a song…"). You won't find many of the same words that make up the most typically played version of "St. James Infirmary" today, but this at least is a back story that makes some of the latter's sentiments perfectly logical: The singer makes a jealousy-tinged boast and turns quickly to thoughts of his own death because his "baby" just died of VD. Dig?

The ballad traveled the world. There is a black West Indian version from the 19th century. And there's one from Kentucky (dated to 1915) that seems to have been adapted to refer to a specific local scandal involving a former policeman involved in a brothel-based slaying that led to his own execution. Another version of the ballad traveled west with pioneers as "The Cowboy's Lament." It's basically the same story again, but the linen-wrapped fellow is a cowboy spied on a Laredo street. ("Get sixteen cowboys to carry my coffin, Get sixteen pretty ladies to bear up my pall…"). Sometimes the request is for a bunch of gamblers to carry the coffin.

Alan Lomax appears on this Folkways disc — singing. He contributes a "Negro version" of the ballad that he and his father collected in 1934 from prisoner in Sugar Land, Texas. It's called "St. James Hospital." Here it's worth noting that up to this point on the disc, none of the versions has the melody of the modern "St. James Infirmary." (It's also worth nothing that Lomax is not much of a singer.) Instead they use the melody of the song we know today as "Streets of Laredo," which has been recorded by Johnny Cash, Marty Robbins, Willie Nelson, Buck Owens, Arlo Guthrie, and many others. Apparently the "Rake" cycle splits in two directions, one leading to Laredo, the other to St. James' Infirmary. The tune Lomax sings links the English folk song to the jazz standard.

* * *

This raised more questions, and trying to answer them has been an interesting — if ultimately frustrating — process. We live in a moment of very intense documentation. Every cultural event — hell, every wedding — is captured on video, in photographs, written up in web logs and emails. The historians of the future will have an embarrassment of riches to work with, no matter how trivial their inquiries may be. And I sometimes wonder if they'll have much left to inquire about, given how few secrets are left in our real-time culture. It's startling to look back less than 100 years in search of answers, only to confront the alien idea of the unknowable.

We know that Irving Mills was born in New York, the son of immigrants from Odessa, Russia. As young men, he and his brother Jack worked as "song pluggers" (promoters), and in about 1920 they set up their own music-publishing firm. At the time, such firms made money by selling sheet music. Live performances and even recordings were basically seen as a way of promoting such sales. Jazz was commercially popular; Mills Music also sold novelty rags and blues. They would buy songs from musician-writers for a flat fee, and own them outright. They once bought all rights to 21 Fats Waller songs for $500.

The forward-looking Mills did a pretty good job getting involved with new technologies like radio, and was apparently a pioneer in sending free recordings to publications to garner publicity. (Recording sales overtook sheet music in the mid 1930s.) He also started working as an agent, most famously for Duke Ellington, under an arrangement that allowed him to take partial writing credit on dozens of early Ellington tunes, many of which he probably did not contribute to at all. For this reason, Mills is generally recalled as a bit of a scoundrel; just about every time I've read some passing mention of him in liner notes or jazz books, it's dismissive at best. There's so much more to say about Mills, but seeing as how he had little to do with New Orleans, I'll move on.

In 1927, the poet Carl Sandburg published a book called American Songbag, a collection of 280 songs (music and lyrics and very short explanatory introductions) from "all regions of America." About 100 of these he describes as "strictly folk songs," never before published. "Though meant to be sung, [the book] can be read as a glorious anthology of the songs that men have sung in the making of America." One of the songs is called "Those Gambler's Blues." Two sets of lyrics are given for the melody, one collected from someone at the University of Alabama, the other given by two sources, one in Los Angeles and one in Fort Worth. There's no mention of a composer, which rather strongly implies that this is one of the folk songs with no known author, which these days we would see credited to "Traditional." The lyrics contain much of what we hear as "St. James Infirmary" today; the melody (I confirmed with a friend who reads music) is basically the same.

The only recording I've been able to find that pre-dates Armstrong's is a performance by Fess Williams and his Royal Flush Orchestra, made February 25, 1927 in New York City. On the CD version, the song is listed as "Gambler's Blues," and, maddeningly, the writer credit is "Moore-Baxter." This might refer to songwriters Fleecie Moore & Danny Baxter, but that's just a guess, and to tell the truth I haven't been able to get a shred of information on this score, and I've never seen a reference anywhere else to "Moore-Baxter" as the composer. Even more maddeningly, I also came across a single stray reference to Don Redman as the song's writer. I don't know what to make of these outliers. They're unsourced. Maybe they are mistakes.

The jazz reference books I've seen that address the question of the songs authorship tend to offer no specific name, but say that it dates back to 1910, or the late 1890s, etc. In other words they don't help.

Again it's worth noting how the world has changed. Can you imagine someone today getting away with taking credit for writing song that had actually been published in a collection — one compiled by a famous poet — two years earlier? Anyway, I don't know where Irving Mills heard the tune. I don't know why he used the name Joe Primrose in claiming it, as he never seems to have used that pseudonym again. I can tell you that the Harlem Hot Chocolates recorded a version in New York in March 1930, with a singer identified as Sunny Smith. This was actually Duke Ellington's band, with Mills, under another pseudonym, on vocals. He's not a great singer, but he's better than Alan Lomax.

* * *

Now, I'm generally skeptical of music writing that focuses on analyzing lyrics, and I deplore attempts to treat lyrics like poetry. However, I am obviously very interested in that one lyrical passage — the one in which the singer suddenly shifts from lamenting his lover's death to bragging that: "She can search this whole world over; she'll never find another man like me."

There's a lot of tweaking and futzing and rearranging of lyrics in various recorded versions of "St. James Infirmary" that I've heard. In the "Rake" songs the singer was a third-party narrator, relating a tale he heard from the stricken man himself. The oldest "Rake" songs basically ignore the woman, who is merely an undifferentiated "flash girl," not the unfortunate protagonist's true love.

This is even true of "Gambler's Blues." In the most prevalent version, the narrator is in a bar and hears the tale of woe from Big Joe McKennedy (or something similar), who is just back from having visited his lover's corpse at St. James' Infirmary. (This is how Eric Burden did it, old school blues poseur that he is, in what I have to admit would be a great rendition if not for the backup singers going "oh-ooh-whoa" over and over.) But this scene gazing at the woman's lifeless body is an addition to the storyline of the "Rake" songs, and suggests that the deceased was, in fact, the singer's true love, or at least main squeeze, not an ill-advised fling. Most of the more modern jazz versions (Armstrong forward) omit the narrator device altogether and make it a first-person story.

That passage I'm so obsessed with does not appear in the old English "Rake" songs, nor is it in either version of the lyrics provided by Sandburg. In one of the lyric sets he offers, the line is replaced with, "There'll never be another like her; there'll never be another for me." This is the way the Hall Johnson Negro Choir did it in December 1931, and it's also the reading that Bobby Bland went with decades later. It's certainly a more traditional and less jarring sentiment. And it's much less interesting.

The line is omitted altogether from Fess Williams' version from 1927 take, which skips straight from the image of the dead woman to the narrator discussing his own funeral. The version that Mills (as Sunny Smith) sang in 1930 basically has it both ways: After seeing his baby on that long white table, he first "wish[es] it was me instead," and then throws in the "search this whole world over" verse right afterward. Another version that Mills was involved with, recorded by Mills Merry Makers in January 1930, has Charlie Teagarden (younger brother of Jack) on vocals, and delivers a take that works so hard to get the verb tense right that it sounds like grammar teacher delivered it: "She could have looked this wide world all over, never she'd never have found a sweet man like me." (Emphasis added.) It's actually a nicely done vocal, but that reading of the line is ridiculous, and completely misses the mysticism and the nastiness of the eternal vengeance implied by saying that even in the afterlife she'll never find such a man. It also waters down the sense that the singer affirming his own life with a kind of proud desperation. Which to me is the whole point.

* * *

In New Orleans, the lyrics are pretty much always performed the way Armstrong did them. The most recent recorded version I know of is on last year's The Marsalis Family, with patriarch Ellis and all four of his musician sons. Harry Connick sings — and uses the lyrics that Armstrong did.

How did the song come to Mills' attention? Did he hear the Armstrong recording? The Fess Williams? Some other recording? Where did Armstrong pick it up? Was it being played in New Orleans when he was growing up, hanging around Storyville? The Teagarden brothers were New Orleanians — is there a clue in their connection to Mills?

I don't know, I don't know. Maybe I never will. Still, that recording of Armstrong delivering what I think of as the key line does not have a precedent that I am aware of. But perhaps this is the way he had heard it performed in New Orleans, before he left for Chicago in 1922. I have no proof of that whatsoever, of course, but I think it is still too soon to say that the song has "no connection with New Orleans whatever." Because every time I hear some local brass band playing the tune, I always say to myself: "No connection with New Orleans? That just can't be right."

June 2003

Thanks to the following individuals for help, feedback, interest, or in several cases the simple willingness to put up with obscure questions from a total stranger: Cynthia Joyce, John Hornsby, Morris Hodara (and here I'd like to plug The Duke Ellington Society, Box 31, Church Street Station, New York City 10008), David Hajdu, John Hasse, Tom Morgan, Gene Anderson, Tom Piazza, Bruce Raeburn, and Michael White.

POSTSCRIPT June 27, 2003:

My initial call for feedback, information, and tips on this story has yielded some interesting results. For example, I have been told of several other versions of the song. One that I'm pretty curious to track down, but haven't yet, is "Touro Infirmary," recorded by Dr. John. Another is by a band called Snakefarm.

One reader noted that near the end of Robert Stone's first novel, A Hall of Mirrors, which happens to be set in New Orleans, the protagonist pays a visit to a morgue in a scene that seems very likely to have been inspired by "St. James Infirmary." So if any of you know Robert Stone, ask him about this.

Finally, yet another reader has directed me to Bob Dylan: Song and Dance Man III, by Michael Gray. Actually this unbelievably generous individual has sent me a chapter of the book that is relevant to my efforts. It concerns the song "Blind Willie McTell," and it introduced to me several important leads, which I am in the process of pursuing now. I won't go on about it here, but there's work to be done that involves more on blues variations on the song, on the historic interplay between black and white folk music, and on African-American cowboys.

In a quick note on the experiment with "viral reporting," I gather the link has been sent around a bit, but most of the response I've gotten has been directly from Letter From New Orleans subscribers. And I'm amused to point out that the best tip, on the book cited above, came from somebody who apparently lives in or near my neighborhood. So there's the power of the Internet for you. Anyway, so far the project does not seem to be spreading quite as quickly as, say, a scatological Flash joke, but I'm preparing to redouble my efforts by bothering more people directly. We'll see what happens -- although I imagine it will be a number of months before I can really revisit this story in any substantial way.

And finally, something else I learned from the early feedback is that I made a mistake. I wrote: "The Teagarden brothers were New Orleanians — is there a clue in their connection to Mills?" Actually, the Teagarden brothers were Texans. So I have had that sentence removed. And I've had a stern talk with my research and fact-checking staff. I also offered my own resignation, but so far I have refused to accept it.

Anyway. I am still anxiously craving more feedback, more information, and more tips. Please send them. Please forward the link to anyone who might know something, who might know someone who might know something, or who might know someone who knows someone who knows something. I want to know. . . . i3

In London there was a St. James's Infirmary at Poland street according to: Boyle's View of London, and its Environs; 1799. London, Printed and Sold by P. BOYLE, At his Court and City Guide Printing Office, Norris Street, Haymarket, and may be had of Richardson, under the Royal Exchange; Lee & Hurst, Paternoster Row, and all the principle Booksellers in the Kingdom.i4

St James' Road Infirmary

In the early 1900s, rapid growth in the population in the Wandsworth area resulted in an increase in the demand for poor relief. Remedies considered included the farming out of poor-relief applicants to other Unions, and the removal of Battersea from the union. Instead, the union purchased the site of the St James' Road Industrial School from the Westminster Union for £21,000. The school was originally erected by St James' parish in around 1852 to house juvenile offenders. By 1884 it could accommodate almost 200 children.

St James' Road Industrial School site, 1888.
 

In 1904-5, the existing buildings were converted to act as a branch of the Garratt Lane workhouse, known as the St James' Road Branch Workhouse. In 1910, a new pavilion-plan infirmary was erected on site, on ground that had previously formed the workhouse gardens. Designed by James S Gibson, the new buildings comprised a central two-storey administration block linked by a long corridor to three-storey ward pavilions at each side, with female wards at the north, and male wards at the south. At the end of each side was a double pavilion containing 204 beds. Half-way along each side was a single pavilion with 96 beds, designed in a way to allow later expansion into a double pavilion. Each pavilion had WCs in projections partway along the block. The ends of the pavilions had large bay windows and sunning balconies. The existing workhouse buildings, which fronted onto St James' Road, were retained alongside the new infirmary.

Wandsworth St James Infirmary site, 1939.
 

Wandsworth St James Infirmary site, 1914.
 

In 1930, the infirmary was taken over as a general hospital by the London County Council. In 1931-5, the original single pavilions were doubled up, and the old workhouse blocks replaced by a new central complex. This contained out-patient, casualty and admissions departments at the west, X-ray and physiotherapy departments at the north, and staff quarters at the south.

The hospital closed in 1988 and the buildings demolished in around 1992. The site has now been redeveloped for housing.i5

While there never was a St. James Infirmary in New Orleans, there was a St. James Hospital. It began life in 1859 as a hotel on Magazine Street between Natchez and Gravier streets. During the Civil War it became a Union military hospital. But in December 1864, rights to the building were returned to the owner, A. B. James, and he returned it to its original use as a hotel.i6