Main Directory Index

Washington Phillips

If you are among the growing number of people who have discovered the recordings of Washington Phillips and would like to know more about his mysterious instrument, you are in the right place.

Yazoo Records made the decision to re-issue the Washington Phillips CD, and is proud to have contributed information regarding Phillips's instrument for the new notes.


Washington Phillips

Washington Phillips at the time of his first recording session in Dallas,
December, 1927
Photo courtesy of Lynn Abbott

In my opinion, the music of Washington Phillips represents the absolute height of rural originality. It would have been unusual enough if he had merely acquired and learned to play a fretless zither, an instrument with virtually no known performance tradition. But it appears that what he did was to re-configure two fretless zithers, to expand the range of both the melody and accompaniment sections, to change the base key by roughly a fourth interval, and then to become a highly skilled player on his creation, producing other-worldy tones unlike those made by any other instrument, including a normally configured fretless zither. And while the country virtuoso deftly tickled celestial sounds out of his instrument, he sang. All this, all by himself and presumably with no help from anyone, in the isolation of rural Texas, sometime in the early years of the 20th century.

For decades, it was believed that Phillips accompanied himself on a dolceola when he recorded, and until recently there was good reason for that belief. A note made by the studio at the time Phillips recorded credited his self-accompaniment to a "dulceola" (as it is commonly misspelled). The records themselves provided no enlightenment at all, noting only "novelty accompaniment" (see below):

There was no known photo of Phillips with his instrument until relatively few years ago, around 1985, when Lynn Abbott of Tulane University found the image shown above. It was in a January, 1928 issue of The Louisiana Weekly, which is archived at Tulane. The context, a Columbia Records press release, is made evident by the caption.

But even after the photo surfaced it wouldn't have been unreasonable to have assumed that Phillips owned a dolceola in addition to the two zithers he is pictured holding. Thus, there was no real reason for anyone to believe that his instrument was other than as stated first-hand by an eyewitness at the time he recorded.

However, in light of recently unearthed evidence, including the photo, it now appears that the studio (specifically, Frank Walker) assigned the "dulceola" identity to Phillips's instrument for lack of knowing what to call it. This isn't surprising; the identities of instruments belonging to the family represented at this site continue to this day to be a source of mass confusion.

So, Walker assigned to Phillips's instrument a trade name for a zither-like instrument with a gizmo attached, but he didn't get this name from the instrument. Here's one theory on how this could have happened. It seems likely that Walker was interested enough in the instrument to have asked some questions about it and that Phillips mentioned that the instrument had previously had a gizmo attached. (As to why he would have said this, read on.) So if we picture Walker in need of a trade name for a zither-like instrument with a gizmo attached, and searching by whatever means for same, it's probable that the dolceola would have been the first to surface. The instrument was promoted by a very active ad campaign during its years of production, which would have ended about 20 years prior to Phillips's first session.

At any rate, it appears that Phillips’s instrument was in fact not a dolceola. Multi-multi-instrumentalist Gregg Miner has devoted much time toward trying to identify the instrument(s) played by Phillips on the recordings, and he has graciously made his findings, theories, and music analyses accessible at his Dolceola Pages. If you're at all interested in the mysterious Washington Phillips, Gregg's pages are required reading. (You'll want to explore the unbelievable assortment of other unusual instruments too, while you're there.)

And for a captivating look at the human interest side of the Phillips story, be sure to read Michael Corcoran's excellent article, which Gregg has also kindly made available at his site. As far as I know, Corcoran was the first to introduce publicly the notion that the instrument heard on Phillips's recordings is probably not a dolceola. His piece provided the inspiration for Gregg's research into the matter, which in turn provided the spark for mine.

As an aside, Corcoran's research turned up another photo of Phillips that's nearly as interesting, in my opinion, as the one with his instruments. This photo is from 1950, some 20 years after his recording days. It pictures a 70-year-old Phillips, this time posed with his mule team. Click here to view it.

So, we all owe a debt of gratitude to Mr. Corcoran for his research and subsequent reporting of non-dolceolan evidence, and I am grateful to Gregg and to Kelly Williams (top authority on fretless zithers and an excellent player of same) for cultivating the idea that the instruments Phillips is pictured holding are the ones he used to accompany himself on his recordings. The photo above is from the exact time Phillips recorded, so it certainly makes sense.

As another aside, on the subject of the photo of Phillips with his instruments, it appeared in the Louisiana Weekly in January, 1928. The accompanying text told of coming new releases by Phillips and other artists. Among these was Willie Johnson, who is also pictured on the same page. He is seated in front of a hood-up grand piano. Johnson lived and died in a state of absolute poverty. This photo was not taken in the parlor of his home, nor is this a barrel-house piano; it's obviously a studio instrument. Both Phillips and Johnson recorded in early December, 1927. The Columbia press release appeared the following month. I feel certain both photos were taken during Phillips's and Johnson's sessions at Columbia's studio in Dallas. I only include this on the odd chance that it could be useful toward someone tracking down a better copy of the photo. Click here for a comparative look at the two photos.

So what are these instruments Phillips is holding? Both are without question products of the Phonoharp Company of East Boston, Mass. As first pointed out by Kelly Williams, the one on the left is a celestaphone, which is a Phonoharp chord-zither of 4/30 configuration, but with a spring hammer gizmo for the melody strings. The gizmo is missing from Phillips's instrument, as is commonly the case. Phonoharp made a standard 4/30 zither without the gizmo, but this instrument and the celestaphone-minus-gizmo are not quite identical. As noted below, there is one slight but potentially significant difference between the two. Incidentally, "4/30" refers to the fact that these instruments have 4 chords and 30 melody strings. The melody strings are arranged in 15 pairs and normally tuned to two diatonic octaves of C.

The instrument on the right appears to be a Phonoharp No. 2 1/4 chord-zither. The No. 2 1/4 was produced in two forms. One has single melody strings, the other has pairs. The latter is the Phonoharp 4/30 model referred to below.

Trying to figure out how Phillips had his instrument set up presented some interesting challenges right off the bat. First, the melody section of the dolceola, fretless zither, and celestaphone are all two octaves of C. Phillips plays every number in a key somewhere around F. But aside from the altered key, the range of Phillips's instrument was an even bigger stumper. The dolceola, fretless zither, and celestaphone all have a prescribed melody range of two octaves. Phillips's melody range spans nearly three. His accompaniment range is out of bounds too. The question is naturally how he accomplished this.

It appears in the photo that the two zithers are joined together to form one instrument. And it appears that the one on the left is strung with only melody strings and that the one on the right is devoted entirely to accompaniment. It looks like the zither on the left has no bass strings; that they have been moved to the zither on the right, which has more than the normal number for that instrument (four). If Phillips's No. 2 1/4 zither was of the variety that has paired melody strings, he would have potentially had at his disposal a total of 92 strings. However, the photo suggests that not nearly all the pins of either zither are strung. But like any observation involving any degree of detail, made through the visual gravel of the old halftone photo, all of the above is open to question pending a clearer copy of the image coming forth.

At first I made no distinction between candidates for the instrument on the left; it was clearly either a Phonoharp 4/30 zither or a celestaphone. I thought the two instruments were identical, other than whether or not they began their lives with the hammer gizmo. But I noticed one possibly significant difference between them. Here is a comparative look at the two.



In both images, the vertical white line with arrow pointing to it indicates the point at which the melody strings end, the far left hand boundary of their range. As can be seen, the melody strings of the celestaphone are spaced considerably wider than those of its gizmo-free counterpart. This is to accommodate the hammer gizmo, so that the hammers are spaced far enough apart to make it playable. On the zither, the junction of the head bridge's two pieces falls right at the end of the melody section (15th pair of melody strings, going right-to-left). The celestaphone hits at the 11th pair, and this is true of no other Phonoharp zither model. Here is a look at the pertinent section of the Phillips photo.



It seems safe to say that the photo positively identifies the instrument in Phillips's right hand to be a celestaphone, by the number of little dark-colored somethings that dot the longer, more steeply slanted section of the bridge. It can be assumed with certainty that these correspond with strings.

Even though the body of the celestaphone is a bit wider than that of the 4/30 zither (by about 1/2"), the accompaniment section gets shorted. In playing the two, the chords of the celestaphone are noticeably crowded and thus considerably more difficult to play. The signifigance of this is that it offers possible motivation for Phillips coming up with the two-zither setup...that his solution to this problem was to add a second zither devoted to accompaniment. Furthermore, the roomy melody string spacing of the celestaphone would be a bonus as well, especially being that they are paired.

I have spent a good deal of time listening very carefully to the recordings of Washington Phillips and in turn experimenting with variously configuring actual Phonoharp instruments of the models Phillips is holding in the photo. I feel that my labors have been fruitful and that the configurations I have arrived at are at least reasonably close approximations. For detailed information on both reconfigurations represented below, see the Configuration Specs Page. I have used one zither rather than two for each example; as mentioned above, I don't believe he used nearly all of either of his. There still is and perhaps always will be some degree of mystery in regard to the exact set-up of Phillips's instrument(s), but in the meantime, I'm satisfied that he played Phonoharp zithers on his recordings, rather than a dolceola. The photo tells us this, and to my ear the recordings do too.

But beyond mere theorizing; here's a chance to listen for yourself to the difference in sound between the two instruments in question. Below are sound clips of the dolceola and both of the instruments Phillips is holding in the photo, a Phonoharp No. 2 1/4 chord-zither (of the 4/30 type with paired melody strings) and a celestaphone. The latter has the gizmo removed like Phillips's, which from the functional standpoint changes it to a chord-zither.

I admit up front that this is an unfair comparison, and for more than one reason. First, the zithers are re-configured to approximate my supposed set-up of Phillips's instrument; the dolceola is not. Second, the bits played on the zithers are from Phillips's songs; the piece played on the dolceola is not. I am not a keyboard player and am only barely familiar with the dolceola; I can't play any of his pieces on it. But for that matter, the performances of the pieces played on the zithers are given to be only rough approximations of Phillips's actual playing; the purpose of the zither clips is solely to illustrate the instruments' sound, as compared to his. And third, the dolceola represented here is the only one I have ever heard, thus I have no idea what they are supposed to sound like. It may be that others sound different than this one. That said, here are a few sound samples for you to compare. First though, here is information on hearing Phillips himself on CD.


Washington Phillips on CD

We are fortunate indeed that fate frowned on the idea of Phillips's music going to the grave with him. He was called upon to record a few sides in 1927-29. And thanks to Yazoo Records, Phillips's recordings are available on CD; on two of them, actually. The recordings were originally released under the title "I Am Born to Preach the Gospel" (Yazoo #2003). The notes of this release were written with the assumption that Phillips was playing a Dolceola. More recently, the recordings were re-mastered and new notes were written which present the instruments actually played by Phillips. This release is entitled "Key to the Kingdom" (Yazoo #2073). By clicking on either of the titles above, you can check out the CDs.

Unfortunately, the sound clips for the newer CD weren't working at the Yazoo site as of last time I checked. If this is still true, you can try at Amazon. In fact, they offer clips of every title on the CD.




First up, here's one played on the dolceola. This piece was recorded years ago as a piano solo.

Cue Ball Blues


Phonoharp 4/30 chord-zither (altered)

Here are a couple of fragments from two of Washington Phillips's songs, played on a specially configured Phonoharp 4/30 chord-zither. I believe this configuration to be similar to how Phillips's instrument was set up at his first recording session, December, 1927.

Mother's Last Word to Her Son

Denomination Blues


celestaphone (altered)

Phillips recorded three sessions, all in the first week of December, in three consecutive years, 1927-29. The configuration of his instrument changed slightly over the course of the time he recorded. Here is the instrumental break from "The Church Needs Good Deacons", played on a specially configured celestaphone. I believe this configuration to be similar to how Phillips's instrument was set up at his third recording session, December, 1929. My thanks to Pat Conte for the session date information.

The Church Needs Good Deacons