The Instruments of Washington Phillips (SEE ALSO MY Washington Phillips CD study)
by Gregg Miner and Kelly Williams, May, 2003
(Last update, September, 2015)

from The Dolceola Pages
created by Gregg Miner, as part of

(To use footnotes, click on a Note, then use Back button on your browser to toggle back to text)

Update, September, 2015: Yet another candidate for Washington Phillips’ instrument has been revealed!  See my blog “What Are They Playing in Heaven Today?” for this new discovery and my analysis on this and all of Phillips’ incredible instruments.

Note: This in-depth study, first published on my web site in 2003, was the first to explore the actual instruments Phillips played on his recordings.  At this point (2015), the world seems to have accepted the fact that the “Dolceola” provenance was simply an unusual error.  Eventually, fretless zither collector and traditional music performer Garry Harrison (1954-2012) expanded on this work, conducting numerous stringing experiments and demonstrating (essentially proving) how Washington Phillips might have achieved the complex results we hear on the records.  I have not updated our original article below, leaving it as a historical record of the events and a useful resource for future researchers.   However, I have followed up on Garry’s work by clarifying a few points directly below.GM

Addendum, September, 2015 by Gregg Miner

Garry’s work, archived on his original web site, recently resurrected by his family, remains incredibly detailed and unavoidably complicated by its very subject.

Don’t be confused by his two pages of “test configurations” – he believed strongly (and convinced me) that Phillips played those two pictured fretless zithers simultaneously.  Whether they were permanently fixed together is probably irrelevant.  Each was completely re-strung from scratch with less strings than it had originally, and in significantly different ways. 

One was custom strung for accompaniment, the other for melody in octave pairs (which was completely unique for these instruments!), and which Phillips could play as an octave pair or either octave string individually.  He apparently added extra bridges to get higher notes than otherwise possible.

Garry’s experiment with (and explanation of) his secondary melody bridges was not related to what Phillips did – it was simply a trick to make the octave strings easier for Garry to play singly (a concept borrowed from the chromatic concert Ukranian bandura).

Garry’s theory makes even more sense in light of the discovery of Phillips’ original “Manzarene” which was two feet by three feet and likely covered with strings.  As it wore or fell apart, he probably realized that he could re-create it more simply by utilizing readily-available discarded fretless zithers (remember that these were sold in the hundreds of thousands).  His three yearly recording sessions could have each used different newly re-constructed zither inventions for all we know.

Using Garry’s work (which he rather hoped someone would have done), it would certainly be possible for a serious Washington Phillips interpreter to construct an improved, modern instrument that duplicates the required string configuration heard on the recordings.  Surprisingly, no one yet has, and I imagine one would take the folk world by storm if they went out performing Wash’s music in such a form!

celestaphone.jpg (24856 bytes)

Celestaphone, with hammer attachment missing. Phonoharp Co. ca. 1912

no2-1-4.jpg (28453 bytes)

Phonoharp #2-1/2, ca. 1910

dolceola_1.jpg (42453 bytes)

Dolceola, 1903-1908
This is the best candidate for the instrument in Phillips right hand. It is a fretless zither of the "chord group" family. Originally having a weighted spring hammer assembly to play the melody strings, this is missing (a not uncommon occurrence due to age and deterioration of the unit). There are a total of 46 strings. The melody strings cover 15 diatonic notes in the key of C, from c' to c'''. The chord groups, from left to right, are C (CGCE), G7 (GGBF), F (FACF), & Am (AACE). Phillips would have had to re-string and re-tune dramatically to achieve the tuning used on his recordings.
This is the best candidate for the instrument in Phillips left hand. It is another common fretless zither of the "chord group" family. There are a total of 31 strings. The notes and pitches are exactly the same as the previous instrument, except that the melody strings are single. Again, if this instrument was used, Phillips would have had to re-string and re-tune dramatically to achieve the tuning used on his recordings.
This is the instrument long thought to have been played by Phillips on his recordings. There are still some who believe it may have been used on some or all of the tracks.

See Fretless Zithers for further explanation of the pertinent terminology and instruments.

celestaphone_compare.jpg (37546 bytes) Phillips' right hand instrument (L),
Celestaphone (R). Included is detail from a poorer quality, but full image, showing the pointed area at the top (cut from the above image).

no2-1-4_compare.jpg (38823 bytes) Phillips' left hand instrument (L), Phonoharp #2-1/4 (R).
Kelly immediately identified both of Phillips' instruments as those made by the Phonoharp Company. Only the Phonoharps have an upper bridge slanting towards the upper left in the bass string area (1) - this section is horizontal in all other companies' similar instruments. A very shallow, distinctive curve (2) at the top is unique to most Phonoharps - other makers use a straight edge here. 

As Kelly further explains, the soundhole on each of these specific Phonoharp models is positioned just crossing the horizontal line drawn at the lowest point (highest melody string) of the upper bridge (3). This matches the soundhole position in the Phillips photo instruments. In all other remotely similar Phonoharp candidates, the entire soundhole is positioned well above this line. 

The positions of the tuning pins in the Phillips photo right hand instrument, at the corner between the melody strings and the chord strings, clearly matches the position of the Celestaphone tuning pins (4). This instrument is double strung in the melody section. One can just make out two rows of tuning pins. Even if you find it too hard to resolve, the position of the clearly visible row of pins between the bridge and the outer edge is that of the outer row of a double row (as on the Celestaphone)(5). Pins in a single row melody instrument are closer to the bridge, further from the edge. 

Other clues: A bare hint of the decal around the soundhole is visible on both instruments. 

Both of Phillips instruments have no end cover. On the Celestaphone, the hammers are integral with the end cover, which is metal. If you discarded the hammer mechanism, you'd have no end cover.

phillips.jpg (45163 bytes) To help verify the models, Kelly Williams approximated a matching self-portrait, holding the two candidate instruments in roughly the same position. To compare: Four distant relatives all independently called " Wash" from 5'8''- 5'9" and "stocky" or about 180 lbs. Note 1 Kelly is nearly 6', 195 lbs. Even so, the two Phonoharps seem to look a little large in Phillips' hands.
kelly.jpg (61095 bytes)
Finally, a Dolceola with the keyboard assembly removed (R), compared to Phillips right hand instrument. Very dissimilar. dolceola_compare.jpg (48374 bytes)

In addition to Kelly's theories and my analysis, Garry Harrison (who more or less agrees) has gone even further and is re-configuring Phonoharps and figuring out how to tune and play them to approximate Phillips' unique playing. He's got two pretty amazing MP3s available on his own special Washington Phillips section. Garry also proposes a new dual-zither theory (completely plausible, as far-fetched as it sounds!).

A Homemade Instrument?

In Michael Corcoran's investigation into the instrument Phillips may have played, he was focused on evidence to verify or disprove the Dolceola as a candidate. In this pursuit, he turned up three eyewitnesses who described a "home-made" instrument. Note 2  He even said that one witness was shown the actual picture of Phillips with the two Phonoharps, but said "that's not it either." Note 3 Corcoran admitted that he wasn't a musician (and not familiar with the instruments of this study), while graciously providing me with phone numbers for some of the key witnesses. After talking with four of them Note 4, it became clear to me that these relatives and neighbors were simply repeating the "common knowledge in the community" - essentially communal folklore - that "Phillips made his 'harp' himself." Corcoran's article(s) leads one to believe that these witnesses had formed the opinion on their own from personal observation, but this was probably not generally the case. Indeed, three could not describe the instrument whatsoever. Note 5 

But one could - and did. This was 85-year-old Earl Phillips, a second cousin (with a PHD in Economics), who definitely recalled a homemade instrument, and gave me some pretty descriptive clues in a phone conversation on April 16, 2003. Earl's testimony was based on numerous occasions watching "Cousin Wash" perform at the three local churches (his was the "African Methodist Church"), and various other functions in Simsboro (Texas), "a mixed community where whites and blacks got along. Whites would often go to black gatherings and sing along to the same songs." Phillips would "play and sing, then preach in any of the churches (white and black) as they asked him." 

I prodded him about as many details of the instrument that he could remember, and I often asked two or three times for verification, as it was hard for Earl, a layperson, to communicate instrument specifics to me over the phone (and vice versa). Nevertheless, his memory seemed consistent, lucid and specific - and adamant about points in which I expressed surprise. The instrument he saw was "oblong, and solid, with no soundhole (like a guitar)." Note 6 It had about "twenty strings - basses on the left, then middle, then high notes." Note 7 He probably played it at an angle Note 8 and only with the right hand - "sometimes on his lap, but usually on a table - where he always held it with his left hand, to keep it from sliding." He played with a thumbpick and probably used all five fingers - the little finger playing high strings. Note 9 When Phillips "wanted to sing loud, he would raise his head and throw back his shoulders.” Therefore, “he’d only look at the (zither) at the start,  then not need to look." The instrument was definitely "made by him (Washington Phillips)," and a neighbor had said, it "looked like part of a piano."

I'm not sure what to make of this. Can these seventy-year-old remembered images be trusted? Note 10  I suppose it's not impossible that what Earl saw was really one of the Phonoharps - perhaps the small 31-string model (missing strings?), weather-beaten so badly that it now looked primitive and "home-made"? But, there is also no reason to think Phillips couldn't have created his own version. In either case, it sounds as if Phillips, in his "retirement," took to playing his "harp" with one hand, which would have simplified things dramatically (though he still may have had all five fingers of that hand blazing away!). Note 11

In conclusion: I don't think we'll ever be able to resolve Earl's or anyone else's' "home-made" conundrum. If true, would it relate to the recordings? Clearly, a one-handed, twenty-string instrument could not have produced the sounds we hear on any of the 16 CD tracks. Furthermore, if we accept the fact that something close to Earl's description was made by hand by Washington Phillips, we might assume that the "handmade" instrument mentioned by the other eyewitnesses was this same one, or something similar. In other words, not an instrument (whether Phonoharp or Dolceola or ?) that Phillips recorded with. The remaining eyewitness, Columbia Records’ recorder Frank Walker, certainly saw the recorded instrument. I'm not sure though what his exact words in the "1961 interview" - Pat Conte has said "Convinced that Phillips employed a home-made instrument, Walker..." [per his liner notes for the Yazoo release], "Walker ...recalled it as a strange 'home-made' looking thing..." per a 1997 Internet bulletin board], and Andy Cohen quoted Walker as "it was some contraption he built himself..." in his 1999 EMI article. As stated on the main page, it's not unreasonable to believe that a thrashed, worn-out, nearly decal-less Phonoharp might appear homemade to a layperson (or Walker). Afterwards, perhaps the two Phonoharps were in such sad shape by the last 1929 sessions that they were discarded - necessitating Phillips' handcrafted version (perhaps utilizing salvaged tuning pins). 

Or perhaps it was just the simplest, least exciting scenario of them all: One surviving beat-up old dime-a-dozen Phonoharp - strung simply and played with one hand. Yet managing to continue to stir up souls and angelic magic for another twenty-odd years. Note 12

Note 1. I got the (very consistent, and independent) descriptions of Washington Phillips from four surviving relatives: Virgil & Jewel Keeton, Earl Phillips, and Wardell Phillips (Wardell is the great nephew of Washington, and now owns the latter's land). Virgil also added that Phillips was "about the same size and color as B.B. King).

Note 2. Corcoran mentions three eyewitnesses: Washington Phillips’ cousin Virgil Keeton, ("harp-like instrument that he made himself") and Phillips’ neighbors Nell Blakely ("a homemade banjo that he laid down flat") and Durden Dixon ("box-like instrument he made himself out of the insides of a piano")

Note 3. Corcoran told me this in our phone conversation in March, 2003. The witness was Virgil Keeton. Unfortunately, when I later spoke to Virgil, he had no recollection of seeing any photos. After discussing with Corcoran again in April, it may have been that Virgil only responded to the image of the Dolceola (reprinted from an ad) in the CD booklet, not the Phonoharps. It would have been a nice bit of "smoking gun," but as Virgil could not recall any of it, we must discount it.

Note 4. Earl Phillips, 85; Virgil, 82 and Jewel Keeton; and Wardell Phillips, 61 - all during April, 2003.

Note 5. Wardell was only 5 or 6 when he saw it. The Keetons saw it often. When Virgil kept calling it a "harp," I finally asked "you mean like an Autoharp?" He not only replied "yeah," but cryptically mentioned it had "numbers on it." I could not convey the concept of decals to him, nor get him to clarify or repeat that tantalizing clue in any way. He seemed to be saying it looked like any other "harp" - as if he was cognizant of typical Autoharps or fretless zithers (probably all the same to him). Again, "made it himself" was not his opinion, but what he had accepted from "everybody." Could he have been referring to one of the Phonoharps? Impossible to say.

Note 6. I was unable to get an answer about the relative volume of it.

Note 7. Knowing that the Phonoharps had either 31 or 46 strings, I was pretty careful about this one. When asked again if it could've had a lot (30 to 50), he replied "Oh no, not more than 20."

Note 8. I say "probably played at an angle," as Earl first said the strings stretched from "left to right on the board," but later said that the basses were "on the left." Here's a good instance where a picture would've been worth a thousand words.

Note 9. He definitely always wore a “clip on his thumb,” and probably didn't use any other fingerpicks. My continual questioning on this would only bring up the fact that he used his little finger for the high notes (not, as I pressed, his ring finger). Harpist and zitherists as a rule only use the thumb and three fingers. If Phillips used all five on the right hand, then in conjunction with the left on his Phonoharp he would've created a fairly rich, intricate sound (as the recordings bear out).

Note 10. Both Virgil Keeton and Earl Phillips thought their "Cousin Wash" recollections were from the ‘30s. Earl said he was "young" and that Wash was "in his ‘60s or ‘70s." In 1935, Washington Phillips would have been 55, Earl 18, and Virgil 15. In 1940, Washington would have been 60, Earl 23, and Virgil, 20, just entering the army.

Note 11. Interesting that Corcoran didn't mention any of his Phillips "strumming action"-mimicking witnesses demonstrating a two-handed technique. Could they all have been remembering and copying the later one-handed technique?

Note 12. Could any of Washington Phillips' instruments have survived? I highly doubt it. If anyone would know, it would be Wardell Phillips, who inherited the land belonging to Phillips. He agreed that the last instrument he remembers probably just disintegrated, or was discarded. But if Michael Corcoran found Phillips' snuff bottle buried in the dirt on the property, then I wonder.....

June, 2005 Update: Yazoo has released newly mastered versions of the Phillips tracks on "Key to the Kingdom" (Yazoo #2073). Pat Contem who wrote the liner notes for the original CD release, "I Am Born to Preach the Gospel" (Yazoo #2003), provided new notes for the new release and paraphrases the conclusions of Garry Harrison, Kelly Williams and myself.


 to Washington Phillips CD Study

 to The Dolceola Pages

[Home] [Intro] [Artist] [Music] [Instruments] [Reviews]
[Appearances] [Place Order] [Contact Gregg] [Links]

Copyright 1998-2016
Miner Music
All Rights Reserved