THIS PAGE HAS MOVED TO:

http://www.minermusic.com/dolceola/phillips_study.htm

Please Re-Bookmark. Thanks!


Washington Phillips CD study (Yazoo Records)

Work in Progress by Gregg Miner
 (additional input from Kelly Williams, Garry Harrison and others listed at page bottom)

from The Dolceola Pages
(maintained by Gregg Miner, as part of www.minermusic.com)

Michael Corcoran's article

Corcoran's revised article, 2-13-03

My Opinion

The Instruments

My Analysis

Track notes

Sessions / Tracks Spreadsheet

String Pitch

True Dolceola Blues on Leadbelly Recordings

For over thirty years, a growing interest has developed in a little-known gospel singer named Washington Phillips, and the strange, even lesser-known instrument which he played – the Dolceola. With the recent appearance of additional specimens of the previously rare instrument turning up, the mythic status of both Phillips and his instrument were perhaps at their peak - when on December 29, 2002, music critic Michael Corcoran, of the Austin Statesman, published an article presenting two startling revelations...

1) The accepted biographical information was that of a different Washington Phillips, and 

2) He probably did not play the Dolceola. 

As a new Dolceola owner myself, I was as surprised as anyone, and resolved to discover the truth of the matter.

It would be unfortunate if the discovery lessened the mystique of either the Dolceola or Washington Phillips. The Dolceola remains a delightful, captivating instrument, with an equally fascinating story of American ingenuity and entrepreneurship. Also still remaining are the yet-to-be-discussed 1944 Leadbelly recordings featuring Paul Howard, an L.A. session keyboardist on “zither” - as it happens, a Dolceola! Meanwhile, Washington Phillips’ mystique and mythological status should only increase. In my opinion, the challenging analysis and study of the instrument (or instruments) he did play, and the unorthodox way he customized it and taught himself to play it, is much more interesting and magical than the original supposition that he played a standard, more easily playable Dolceola (however uncommon).


My Opinion

Bottom line: On his 16 extant recorded tracks, Washington Phillips is not playing a Dolceola, but a common fretless zither (most likely one or both of the Phonoharps in the photo), albeit with a unique, self-invented stringing and tuning pattern.

Michael Corcoran's article is full of fascinating but contradictory clues about the true identity of Phillips' instrument. Alone, none of them provide irrefutable proof, but they do provide support for a fretless zither theory. I believe that Phillips may have indeed built a "home-made" instrument - but if so, it was probably a later, much cruder and simpler instrument than those played on his recordings. He could even have flirted at some point with a Dolceola (or similar instrument) - but the proof for me is what's on the recordings, and my own ears.

After Corcoran's article came out, I re-visited the Phillips CD, and then obtained and listened to Andy Cohen’s modern recording of two "working" Dolceolas. I can easily hear a night and day difference between the instruments and sounds. Both are pretty, and are often intricate or lush – but in completely different ways. Cohen does things Phillips could only dream of, while Phillips can (and does) play literally a dozen or more overlapping, ringing notes. The Dolceola, struck near the ends of the strings with tiny wooden hammers, is much "dryer" and "brittle" – the Phonoharps, finger-picked more towards the center are less so. My opinion is that it is impossible to duplicate or approach the sounds Phillips makes on a Dolceola. Happily for the legend, it is also all-but-impossible for any of us to duplicate his self-styled virtuoso techniques on a Phonoharp-type zither!

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


My Analysis

My research consisted of two parts: collating and weighing the published documentation, with personal verification of a couple of the key "eyewitnesses," and CD analysis.

Part 1. A recap of the evidence and support for and against the Dolceola theory with my notes following (To use footnotes, click on a Note, then use Back button on your browser to toggle back to text):

For the Dolceola:

  1. 1927: Frank Walker, Columbia Records’ field recorder, writes "novelty Acc., dolceola" in session files. Note 1
  2. ?-1991: Pat Conte provides the liner notes to Yazoo's Phillips CD "I Am Born To Preach the Gospel," providing much information on the sessions and the Dolceola. Note 2
  3. 1970 to 2002: Musicians from Ry Cooder to Andy Cohen record or perform Phillips’ music with Dolceola.
  4. 1991: At the International Conference of African American Music and Literature in Belgium, Dutch musicologist Guido van Rijn argues for the Dolceola. Note 4
  5. 1999: Andy Cohen's comprehensive article in the Experimental Musical Instruments journal. In it, he speculates how Phillips may have retuned the Dolceola and how it was recorded. Note 5
  6. 2002: Michael Corcoran’s personal interviews: Memphis musician Rick Field quoted "When it comes to trusting eyewitnesses or my ear, I'll stick with the Dolceola [theory] until there's proof." Jim Dickinson, who played Dolceola on Ry Cooder’s Crossroads soundtrack, states "I’m 100% sure it’s a Dolceola" and gives examples of "sounds like part of it is going backwards" (?) and "you can hear the hammer action in a couple places." Note 6

Against the Dolceola

  1. 1983: The infamous 1928 photo of Phillips holding what appears to be two Phonoharp fretless zithers is discovered. Note 1
  2. 1999: William Hettrick, the great-nephew of the co-inventor of the Dolceola, himself implies doubt (due to chord/note discrepancies) in his definitive article published in the journal of the American Musical Instrument Society. Note 2
  3. 2002: Testimony of Corcoran’s eye-witnesses. Cousin Virgil Keeton was shown a picture of a Dolceola and stated it was not the instrument Phillips played. Note 3
  4. 2002: Corcoran: At least five eyewitnesses described Phillips instrument as "strummed" or "plucked" with the thumb.  Note 4
  5. 1961-2003: Five eyewitnesses described the instrument as home-made.  Note 5
  6. Corcoran: Yazoo honcho Richard Nevins himself asks, "My ears are telling me that he's plucking the strings. It sounds like a zither to me." Note 6
  7. January, 2004: I finally heard for myself the 1944 Leadbelly sessions, which feature a documented Dolceola (and now you can too!). Just like the instrument should sound - much like Andy Cohen, and completely different from Phillips. 

Skip Notes


Notes

For:

Note 1. Assuming now that it wasn't a Dolceola, I have no idea where Walker could have come up with the term. According to Conte in the Yazoo CD liner notes, Walker was "convinced that Phillips employed a home-made instrument," So why then, did he originally enter “novelty acc., dolceola." into the files? If it was homemade, why dolceola? Could it have reminded Walker of a Dolceola he had seen (in person or from advertisements)? One wonders if Walker (“sensitive to rural music” and from a farm background) was familiar with the hundreds of models of fretless zithers in circulation, sold by the tens of thousands between 1894 and 1927 (when he made his session notes).

Note 2. I received this very thoughtful and generous letter from Pat (whom Andy Cohen calls "a genius," and I found a very down-to-earth, self-deprecating fellow). He explains, 

"Yes, I was in touch with the writer (Corcoran) and followed his research along in Texas regarding the family's thoughts on dolceolas and some kind of plucked invention (I imagine very like a bandura) but I am not convinced there is one instrument system throughout, as I do hear key action on several 78s and plucking arpeggios on others. The lack of any sour notes or mis-struck adjoining strings does lend good argument that there is a string chord group being fingerpicked. Many of us record collectors have argued over this for years with 78s in hand ready to go one way or the other, but whatever his "invention" may be I am almost convinced that one permutation of it seems to be a dolceola re-set up for the left hand work, with the key transport removed, and plucked. A newspaper photo showing him with two zither like things seems to confirm one is a dolceola body. One important disadvantage to listeners of his music unfamiliar with the actual 78s themselves is that sometime reissues cut off the final extraneous noises at the end of a performance, and I recall at least one with key action sounding. So my feeling seems to be at least two different instruments were used over the course of his sessions and one is really a dolceola. The other has to be plucked just for the sheer magnitude of perfectly formed chords and arpeggios that sound seamlessly without lever action being audible."

I would have to respectfully disagree with Pat on a couple things. First, I do hear tonal differences (miking and/or change of instrument) and possibly stringing/tuning differences (again, 2 instruments) in the different sessions - but, I'm sorry - I hear no key action anywhere. Second, I considered the possibility of Phillips using a Dolceola with the keyboard removed (indeed, one of the Phonoharp instruments may very well be a Celestaphone with its hammer mechanism removed) - but I see no need to go looking for another candidate instrument, when he's already shown holding them. I'm not sure why Pat thinks one of the Phillips photo zithers is a Dolceola body - Pat was one of the first to own and play one, and can surely recognize one (it took collector and expert Kelly Williams to inform me which brand and models these were, though the precise models may not be correct [but close enough for discussion]). In Pat's defense, the two instruments do seem to appear larger than normal in the Phillips photo (see The Instruments of Washington Phillips ). Third, I'm curious about the "extraneous noises" at the end of Phillips performances - why would there be key/hammer noise after he's played, rather than during? Admittedly, I haven't heard them, and my ears may not be perfect.

Note 4. I subsequently learned that Guido van Rijn first reissued the Phillips 78's as Denomination Blues on vinyl. He also wrote a paper on Phillips' recording sessions, which might help shed some light on my Pitch/Session study below (however, his paper is based on the Dolceola theory). I have so far been unable to obtain a copy of his invaluable research - this site is where Guido sent me: http://www.bluesworld.com/Sacre.html

Note 5. Andy's re-tuning of the instrument is quite ingenious (just listen to his CD), and I hope to perhaps explore it one day. I haven't analyzed it fully, but don't believe it covers many of the notes I hear on the Phillips CD.

Note 6. I don't know either Rick or Jim, nor understand the "going backwards" comment. Assuming their ears are as good or better than mine, here we have a perfect example of the nature of sound and interpretation. As it turns out, I’m 100% sure it’s not a Dolceola.

Against:

Note 1. I am embarrassed beyond belief that I, myself, saw this picture on the CD cover - but didn't pay it much attention.  I would like to think that, had I had more interest in the recording, it would've dawned on me...why is holding a couple of typical fretless zithers if he's playing a Dolceola? Was his Dolceola in the shop? Did someone hand these to him and say, "Here, just hold these..."? If not a "smoking gun," then certainly a pretty generous clue for all of us! Since the picture coincides date-wise with the sessions, there is no reason not to believe that these two Phonoharps (specially-strung, remember) were the instruments used (either of them, or both). It would be of immense value if a better reproduction of this image were available for analysis. I'm begging anyone who has any knowledge of the whereabouts of the original (or good negative) to come forward.

Note 2. I obtained this article in February, 2003. To me, Hettrick's logic (and detail) leaves little doubt about the unlikelihood of a Dolceola - though in print (and even in personal correspondence) he politely remains neutral.

Note 3. Inconclusive, as Virgil first saw Phillips several years after the sessions were completed (see Notes 4 & 5). 

Note 4. According to Corcoran's article, the three eyewitnesses are Cousin Virgil Keeton ("demonstrated, with a thumb-plucking motion, how Phillips played the strings on his instrument"), second cousin Earl Phillips ("strings that he strummed"), and neighbor Durden Dixon ("...Durden Dixon, said Phillips strummed a..."). Corcoran also informed me after the articles were published that "there were more than three eyewitnesses from the Simsboro and Teague area who recollected Phillips strumming. Cleo Phillips of Oklahoma was one. Doris Neely another. A couple other eyewitnesses whose names I didn't use because it would be kind of redundant." Pretty compelling, as there are several of them, and these are good layperson's descriptions of typical fretless zither technique. However, it's not impossible that what they saw was not one of his session instruments (see Note 5 and The Instruments of Washington Phillips).

Note 5. Corcoran mentions three eyewitnesses: 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"). The 4th is Columbia Records’ recorder Frank Walker (this apparently from a 1961 interview with Walker. I'd like to see the original as 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). The 5th is Phillips' 85-year-old second cousin Earl Phillips, who provided many intriguing clues in a thirty minute phone conversation I had with him on April 1, 2003 (I have Michael Corcoran to thank for his leg-work and providing several phone numbers). However, even with five witnesses calling the instrument "home-made" or "made by Phillips himself," I wouldn't necessarily expect all these laypeople to be able to look at a beat-up old Phonoharp (almost no remnants of decals can be seen in the photo) and necessarily realize it was originally a production instrument. Additionally, after talking to Earl Phillips, Virgil and Jewel Keeton, and Wardell Phillips during April, 2003, I got the distinct impression (from all but Earl) that it was simply "common knowledge in the community" or, rather, communal folklore, that "Phillips made his 'harp' himself." It was not that these witnesses had formed the opinion on their own from personal observation. This theory is further explored in The Instruments of Washington Phillips.  It certainly is not impossible to believe that Phillips could have borrowed the Phonoharps' layout (and perhaps even cannibalized zither pins and wood) to build his own version when his instruments fell apart (as they are wont to do). But, be that as it may, since the picture coincides date-wise with the sessions, it seems most likely that the Phonoharps (specially-strung, remember) were used.

Note 6. In a personal conversation with Nevins, he said pretty much the same thing to me. He added that though he hired Pat Conte (the best expert available) to write the liner notes, and let them stand, he was never convinced. The issue doesn't seem of much importance to him, at any rate. 


So what do we glean from the previous? Plenty of circumstantial evidence, but unfortunately no "hard" evidence. Thus the best research tools are our own ears and familiarity with the various types of candidate instruments - the sound and techniques of the Dolceola and the many variations of Fretless ("Chord" or "Guitar") Zithers.

Part 2. CD and instrument analysis:

I admit (and regret) that I was unable to resolve to completion the many strange and wonderful things coming out of Phillips' recordings. I almost immediately recognized the sounds as those of a common fretless zither (and yes, I should have caught this on my first listen. Shame on me). Others besides myself are now catching the obvious glissandos (Track 9 opening, Track 11 end) - common to harps and psaltery-style zithers. However, I also quickly noticed the lack of the standard chord group zither's four (or more) basic 4-note chords. Instead, the I & V chords had too many notes, while the remaining chords had too few, if any. While I had Kelly Williams working on identification of the models (and original stringing of same), I tried to pick out the dozens of individual notes. This was done at several speeds on a digital recorder, much of the time slowed down and lowered to the pitch of Bb (from relative F/F#). At this speed, my ear was still fooled multiple times, and it seemed that my deductions went around in circles. Take Track 14 for instance - which has the cleanest, oft-repeating instrumental sections. Phillips is double-striking the melody much of the time. But is it unison or octaves? OK, octaves - some of them going high-to-low, some low-to-high. But are they played in succession by "slurring" on an octave-tuned pair? Or are they played with the common harp "octaves" technique, with thumb and ring finger on single strings an octave apart? I don't know. Various tracks are possibly the latter, this track seems to use pairs tuned in octaves to my ear. 

For now, I've had to give myself a "time out," and just post my jumbled notes as best I can. These follow below - various spreadsheets in which to analyze the many factors involved, and pitch charts to illustrate the notes I may (or may not) be hearing. Note that I haven't yet entered the actual recording session information - my "sessions" are by instrument pitch - thought to perhaps help differentiate instruments and/or tunings (if any differences exist). These studies are woefully incomplete, and I encourage anyone interested to pitch in any time they want! 

Someone who took me up on my offer is Garry Harrison, who on his site provides more details on the instruments and recordings, and proposes a new dual-zither theory (and completely plausible, as far-fetched as it sounds!). He has now re-configured a Phonoharp or two and provides fascinating and well-executed downloadable MP3s of very close approximations of two of Phillips' pieces. 

By contrast, listen to Andy Cohen's distinctive (and representative) Dolceola recordings. Even more telling - L.A. union pianist Paul Howard also played the Dolceola on eight 1944 Leadbelly tracks (you can now download an MP3 of sample riffs). I only learned in 2003 that the "zither" accompaniment on this session was actually a Dolceola - which, by the way, is a "zither" - see "A piano? A zither? or what?".
After all this - put your Phillips CD back on and see if you now hear a near-virtuoso on the Phonoharp! 


Track Notes

First, a summary of specific examples that, to me, eliminate the Dolceola as an option, while suggesting a fretless zither. For the purposes of discussion, I will call this zither a Phonoharp, which is believed to be the specific model/form of the fretless zithers in the photo.

Notes on the following spreadsheet: I added separate columns to examine:

Note: Phillips only has bass strings for I and V chords, plus 1 A for III chord. 
III major chord uses A bass and an out-of-tune, "dead"-sounding C#, plus melody strings. VI minor chord uses melody strings.

And here are the actual recording dates, courtesy of Pat Conte and Garry Harrison. The spreadsheet lists each track by order of the 5 sessions.
Session 1: December 2, 1927
Session 2: December 5, 1927
Session 3: December 4, 1928
Session 4: December 5, 1928
Session 5: December 2, 1929

Sessions / Tracks

Track

Key

Hz

Session

Chords used

IV chord

C#?

Octaves?

Notes

Non-Dolceola "smoking guns"

4

F#, flat

F#, A = 436

1

I, IV, V

"left hand" chord is high, higher than D at right. Not as "jangly" out-of-tune

1st we hear this: flat but consistent. Goes into mid-range D. D probably with right hand melody bank. Lowest note we ever hear on IV chord (even though other tunes have a lower Bb).

Octaves in bass? Melody is single

.

.

6

midway between F & F#

F#, A = 432.5

1

I, IV, V, IIImaj, VImin

as jangly?

still flat.

Octaves in bass

1st we hear III & VI chords. We only hear root and 3rd of VI min chord (melody strings)

.

9

F#, flat

F#, A = 435.5

1

I, IV, V

very quiet and jangly.

.

Instrumental sections: single notes, octaves played on single strings (thumb and ring finger)

All verses: a nice single string low melody: F, E-D-C. Later, a single string "answer" an octave higher.

unrelated clack at end

11

F#, flat

F#, A = 435.5

1

.

Same III chord with very flat, dead C# as Track 6? III chord uses low E(A?) (below C & G of chords). No chord with E bass, just weird out-of tune half step

Can one hand play Bb> and still play high chimey notes? (octaves?) But intro ends with single notes.

LOTS of notes. While left hand clearly works it's own string bank, melody goes down to Bb (lowest yet).

Glissando at end - 1st and end note are clearly F-f (full octave), and sound like single strings.

7

F, sharp

F, A = 446

2

I, IV

?

very flat. consistent? again into low D

Sounds like high melody octaves

As track 4, but new session

.

8

F, sharp

F, A = 446

2

I, IV

same

very flat. Also an accidental c''# (octave higher)!?

.

.

1

F, flat

F, A = 433.5

3

I, IV

no "jangle"

.

.

.

.

14

F, flat

F, A = 435

3

.

.

C# definitely heard in octaves!

Octaves all over the place - bass/chords, melody. It's unlikely that all this could be played "octave style" on single strings 8 strings apart.

.

Octaves. Two-in-a-row C's in the melody are clearly played on 2 different C strings (either within octave set or melody bank/ then chord bank).

10

F, flattest yet

F, A = 431

4

I, IV, V

.

Right hand seems to be playing high melody while left "appears" to be playing C#/F-D/F. C# finally tuned pretty accurately (Note new session). Consistent pitch?

Full riff: F-E-D-C-D-C-Bb-A-G is in octaves, but could be octave plucked. Check the solo's sloppy parts for octaves separating. Very end has octave melody all down an octave, but again may be alternated. Last high arpeggio - octaves?

the 3 steps sound louder and "over" left hand chords. Which plays the C note? Higher basses?

.

12

F, flattest yet

F, A = 431

4

I, IV, V

.

.

.

.

.

2

midway between F & F#

F#, A = 429

5

I, IV, V

jangle

.

yes, but staggered (singles played with thumb and ring finger for example).

.

What are doubled notes? Doubled high C against melody C or octaves.

3

midway between F & F#

F#, A = 429

5

I, IV, V

?

.

Some bass strings almost certainly in octave pairs. Melody also.

does C-D-C figure on F chord (probably melody strings) - sometimes in octaves.

.

5

midway between F & F#

F#, A = 429.5

5

I, IV

?

.

.

.

.

13

midway between F & F#

F#, A = 430

5

I only

.

.

.

Kelly hears C-D-C-Bb-A. I think the Bb is "suggested", but may be heard at very end? If this riff same octave as Track 11, then might be new melody range. But I think A & C of riff could be simply those in I chord. D from melody bank. If Bb (track 11 for sure), why does he NEVER play it (or low D & F) for IV chord?!

.

15

midway between F & F#

F#, A = 430

5

I only

.

.

.

.

.

16

midway between F & F#

F#, A = 430

5

I only

.

.

.

.

.


String Pitch

Phillips' Fretless Zither Melody String Pitches
(I don't address the issue of octave tuning of certain pairs of strings on the double-course Phonoharp[s]) 


Phillips' Fretless Zither Bass/Chord String Pitches
(Again, difficult to resolve when he's playing single bass and rhythm notes, single notes played in octaves, or octave pairs with one stoke)


Phillips' Fretless Zither Compared to Dolceola
Note that the candidate models of Phonoharp (single or double strung) contain a range of c' to c"'. Chords would originally be in groups of only 4 strings, 4 or more chords available - but Phillips probably changes this.
Bill Hettrick explains in his AMIS Journal study the mis-match of chords on the Dolceola and Phillips recordings. No need to elaborate.


Special Thanks to: Andy Cohen, Pat Conte, Michael Corcoran, Jim Garber, Garry Harrison, William Hettrick, Earl Phillips, Kelly Williams.


Copyright 2003, 2004
Miner Music