A Linguistic Analysis of Some West Virginia Petroglyphs

By Monroe Oppenheimer and Willard Wirtz

The West Virginia Archeologist Volume 41, Number 1, Spring 1989

© 1989 WV Archeology Society, used with permission

The March, 1983, issue of Wonderful West Virginia contained a fascinating account of the "Wyoming and Boone County petroglyphs" (Pyle, 1983; Gallagher, 1983; Hyde, 1983; Fell, 1983). Markings on the stone walls of two rock shelters in the southern part of the state had been interpreted by Dr. Barry Fell, described as an emeritus professor at Harvard and "America's leading ancient language expert," as being in old Irish script called Celtic Ogam (or Ogham) and as dating back to the 6th or 8th Century A.D. They were interpreted by Dr. Fell as carrying, in extensive detail, the story of Christ's nativity.

The momentous implications of the reported discovery were obvious. It meant that somebody had come to this country from Europe, carrying Christianity's message, seven or eight centuries before Columbus.

Recognizing that some pieces of the picture needed further development, the 1983 Wonderful West Virginia articles promised that additional information would follow. Three years later, a brief follow-up story in the June, 1986, issue reported that the "Petroglyph Case Remains Open." This second article included none of the further information that the first report had referred to. It recognized that the earlier account had created "considerable controversy," and said that "Assembled Evidence Strengthens Petroglyph Interpretation Case," but left the reader to decide for himself if the evidence is plausible, regardless of the dissension of professionals and avocational archaeologists in the state.

The meaningful use of such freedom obviously depends on having, with respect to this controversial issue, all of the available evidence. This means giving to the public some information that has not so far been published in Wonderful West Virginia.

Perhaps the most critical piece of this information is that by using the "decipherment" methods Fell sets out in his March 1983 article it is possible to find in these rock wall markings not only the nativity story but any other preconceived text one might choose.

It is equally sobering to discover on investigation that Barry Fell's connection with Harvard is as a retired professor of marine biology. His profession had nothing to do with archaeology or linguistics or ancient inscriptions, and he is in fact an extremely controversial figure whose previous decipherments of this same kind have been seriously challenged, after careful study, by American, Irish, English and Scottish archaeologists and linguists. Reviewing the Wonderful West Virginia report of evidence regarding the Wyoming and Boone County petroglyphs, these authorities conclude, spelling out their reasons, that this is a transparent hoax.

Alphabet from Fell's 1983 article
Figure A from Fell's 1983 article
© 1983 WV Division of Natural Resources, used with permission
With respect to the first of these points, Fell provides the careful reader of his 1983 Wonderful West Virginia article with the basis for full understanding of his "decipherment" process. He starts by setting out the Ogam alphabet, a series of perpendicular lines and notches cut along a horizontal "stem line'' which is Ogam's critical distinguishing characteristic (Fell, 1983: 12 Figure A).

Cover illustration, the Ogam alphabet
Ogam consonants were formed by cutting from one to five short perpendicular lines above or below or across the stem line. One perpendicular line cut below the stem line ( ) was, for example, the letter "b". One line crossing the stem line ( ) was "m"; two crossing lines ( ) were "g"; three crossing lines ( ) were "n"; and five across ( ) made "r".

Vowels were made in Ogam by cutting notches in the stem line; one notch the letter "a", four for ''e", five or "i", two for "o,", three for "u" (see cover for Ogam alphabet).

Luther Elkins Petroglyph
Figure 1. Luther Elkins Petroglyph Site, Wyoming County.
[Chalked by Robert Pyle. No one should ever chalk a petroglyph: it interferes with possible chemical analysis, and fixes the chalker's biases and preconceptions as to what is legitimately part of the figure -- and what is not]
© 1983 WVDNR, used with permission
Turning to the slashes on the Wyoming and Boone County rock walls, it is possible to find in them a rough suggestion of some Ogam characteristics (Figure 1).

A good many of the slashes are perpendicular; but they are cut so closely together, twenty to thirty in a kind of a row, that separating them into the groupings of from one to five lines that the Ogam alphabet requires appears impossible. Three or four approximately horizontal lines cut across about a third of the perpendicular lines, with two or three horizontal lines intersecting the same perpendicular line; the remaining perpendicular lines are unconnected with any horizontal line at all. There are no notches (which would signify vowels ) in any line or any place on these petroglyphs.

The carvings include several arrangements of lines that Fell calls "rebuses." Among these, one set of eighteen short lines project out from a center in a circle - like the petals of a daisy, the spokes of a wheel, or with a little more imagination the rays of the sun.

Fell's interpretive method applied to the portion in the photograph
Fell's interpretive method applied to the portion in the photograph
© 1983 WVDNR, used with permission
Part of Horse Creek petroglyph
Part of Horse Creek petroglyph, Boone County (unchalked). Compare closely with Fell's drawing.
© 1983 WVDNR, used with permission
Fell (1983: 15-19) describes in a series of sketches the steps he takes to transform these slashes and "rebus" figures into Ogam consonants. But no vowels. First, recognizing that horizontal stem lines are essential to Ogam script and that the West Virginia markings are notably short of lines that could even conceivably serve as stem-lines, Fell simply adds whatever stem lines he chooses at whatever places he decides on. Where a set of perpendicular lines is intersected by two or more of the few horizontal lines in the carvings, he makes an arbitrary decision about which of these horizontal lines to use.

This results in the perpendicular slashes on the cave walls being given whatever "meaning" follows from Fell's arbitrary construction of stem-lines. Taking, for example, three perpendicular lines ( ), he makes them into an "f" if he constructs his stem-line above them ( ), into a ''t'' if he places the stem-line below them ( ), or into an "n" if he runs the stem-line across them ( ).

Fell next addresses the fact that identifying Ogam consonants depends entirely on the grouping of perpendicular lines; yet those on the walls of the West Virginia rock shelter are not grouped at all. Fell groups them arbitrarily. Taking, for example, five of these lines which cross his constructed stem-line, Fell makes his own decision whether to group these five together, which would give him the letter "r" or to sub-group them into first two and then three ( ) which would mean a "g" followed by an "n"; or to divide them three and then two ( ) which would produce "n" followed by "g".

Fell's "variant forms encountered in West Virginia" (From his Fig. A)
© 1983 WVDNR, used with permission
If no perpendicular lines meet his particular needs, Fell straightens other slashes up. The letter "t", for example, is in the Ogam alphabet as Fell sets it out three parallel lines proceeding up from the stem-line ( ). Needing "t", Fell "finds" it in these figures on the walls: ( ), ( ), and ( ).

Having used enough of the slashes on the rock walls to give him his nativity text, but with a good many slashes left over, most of them running indiscriminately at various angles, Fell announces that these were added to the petroglyph later by some other scribe who used a combination of "Libyan and Algonquian" scripts. These, Fell reports, are to be described in a subsequent article. It has never appeared.

Fell explains the three non-Ogam characteristics of the lines of the West Virginia rock walls - the absence of any except a few horizontal lines, the ungrouped perpendicular lines, and the lines running at an angle - as "variant forms (of Ogam script) encountered in West Virginia (Fell, 1983: 12). Yet these "variations" appear in no verifiable Ogam inscriptions that we have been able to find. Nor does Fell point to any.

He was left with the problem that the liberties he had taken with respect to the lines on the rock walls - adding those he needed, straightening others up, grouping them arbitrarily -gave him a series of consonants but no vowels. In Ogam script, as Fell himself describes it, the vowels are made by cutting notches on the stem line. There are no notches on either the Wyoming or the Boone County rock walls.

This proves no problem, though, to Fell. He announces that these particular petroglyphs are in "the earliest type of Ogam, (which) used no vowel signs (Fell, 1983: 12: Figure A)."

This strains credulity past the breaking point. It isn't just that Fell offers only the flimsiest basis for his conclusion that an early form of Ogam was what linguists call "consaine," a form of writing, that is, which includes only consonants. He ignores, in making this announcement, the fact that the Irish language has been authoritatively identified as not of a nature consistent with the use of a consaine alphabet (Carey, personal communication, 1987, and Fitzhugh, 1978: 167).

What is critical here is that by this device Fell gives himself virtually complete flexibility in "interpreting" any series of consonants he has constructed. If he were dealing only in English he could, for example, make the consonants NGDWTRST mean -if this were the text he had started with - IN GOD WE TRUST. Yet, they could equally as well be interpreted as NO GOOD WATER SITE.

So, it becomes apparent on the face of Fell's own account that what he did here was to work back from his preconstructed text of the nativity story (in English), through the old Irish language, to the markings on the walls of the West Virginia rock shelters. He found whatever "Ogam" consonants he needed by manipulating the slashes as he chose, and then added whatever vowels the test called for. The same thing could be done in "deciphering" a handful of jack straws thrown out on the table by a child. It would be a harmless and mildly interesting game if it weren't played for such high stakes in terms of public deceit.

The conclusion that such a deception has been practiced on Wonderful West Virginia 's editors and readers might appear both suspect and presumptuous if Barry Fell's professional standing were indeed that of "America's leading ancient language expert." The truth is that he falls lamentably short of having earned any such description.

It turns out on investigation that Fell's "analysis" of the West Virginia petroglyphs is one more incident in a practice he has followed over the past fifteen or twenty years. The 1983 Wonderful West Virginia article mentions in a footnote his publication in 1976 of a volume entitled America B.C. Its thesis is that Irish, Iberian, Libyan, and Egyptian explorers came to this country some 2,000 to 2,500 years ago. A number of other alleged discoveries like those in the Wyoming and Boone County rock shelters are relied on to bolster and buttress this theory.

What the Wonderful West Virginia article leaves out is that the Fell book has received widespread and illuminating professional attention. The consensus of recognized archaeologists and linguists is that America B.C. is a complete fabrication.

A professional conference was convened in 1977 at Castleton College in Vermont to consider the "evidence" reported in America B.C. Fell was one of the conference participants. A verbatim transcript of the proceedings was kept (Cook, 1987: 85-96). After others of those present had expressed doubts about America B.C. findings, Fell responded - with vehement invective. He charged his critics with being too "damn lazy" to read what he had written, so "ignorant" that they "can't even hold a Phoenician inscription the correct way up," and with being united in a jealous desire to protect their professions' conventional wisdom against the conflicting theories he has been developing. Displaying pictures of petroglyphs he claimed to have found in American caves and deciphered in Celtic Ogam, Fell refused to disclose their location. "As long as I am an unpronounceable person, I am not going to say where they are.

A 1977 review of America B.C. in the New York Times Book Review, by Glyn Daniel, Professor of Archaeology at the University of Cambridge, England, described Fell's contention as "ignorant rubbish," reflecting a "set attitude of mind" that is "almost indistinguishable from a delusion" (Daniel, 1977). In characteristic fashion, the Smithsonian Institute in Washington published in 1978 a calm, reserved, meticulously careful Statement regarding America B.C. Authors Dr. Ives Goddard and Dr. William W. Fitzhugh, of the Institution's Department of Anthropology, listed five different sets of basic factual errors and anachronisms in the controversial volume that make it totally incredible (Goddard and Fitzhugh, 1978).

Professor F. H. Wilhelm Nicolaisen of the State University of New York at Binghamton, recognized expert on place names, has analyzed Fell's claims in America B.C. that the names of various New England towns and rivers are of Irish Celtic origin. Nicolaisen traces these names meticulously and irrefutably back to American Algonquin derivations (Nicolaisen, personal communication).

During the past five years, a number of widely recognized and respected archaeologists and linguists have had an opportunity to review the 1983 Wonderful West Virginia report of the Wyoming and Boone County petroglyphs. So far as appears, they all reject, on what seems to be solid ground, the Fell decipherment and interpretation.

Among these commentators, Dr. John Carey of the Department of Celtic Languages and Literature at Harvard University, analyzes a series of questions of grammar and spelling that arise in connection with Fell's purported translation of the West Virginia petroglyphs. He points out that Fell "finds" in them sentence structures and spelling that in fact developed in the Irish language many centuries later than the time - the 6th to 8th centuries A.D. - to which Fell attributes these wall carvings (Carey, personal communication, 1987).

The same incriminating anachronism is identified by Professor Proiseas Ni Chathain, of the Department of Early Irish at University College, Dublin. "At its simplest it is impossible to equate proto-Irish forms and primitive Irish with late medieval and modem Gaelic as (Fell) does. In some cases there would be a gap of sixteen hundred years and changes in the language are traceable and have been well documented throughout this period. Dr. Fell is apparently quite unaware of this and no trace of it appears in his published work.... There appears to be no scientific basis for his opinion in respect of his readings of the alleged Ogham inscriptions in West Virginia (Chathain, personal communication, 1983)."

Dr. Cathain O'Dochartaigh, of the Dublin Irish Language Institution, after analyzing the Fell article in the March 1983 issue of Wonderful West Virginia, advises that he does "not think that these rock scratchings can in any way be regarded as representations of Ogam writing." Referring to "what Dr. Fell considers a transcript of the daisy wheel symbol," O'Dochartaigh writes: "Two questions arise. Why did he choose a medial, rather than a bottom, stem line and why this particular segmental division when all one has is a set of eighteen undifferentiated strokes?" (O'Dochartaigh, personal communication letter to Joseph J. Snyder, Harper's Ferry, 1983).

Harvard's Dr. John Carey also comments particularly on the lack of any basis for Fell's treatment of Celtic Ogam as a consonantal alphabet, including no vowels, and on the unreliability of any interpretation based on this theory. He observed that writing Irish or some version of it in this manner "makes it so ambiguous that one can extract almost any meaning one pleases for the resulting string of letters .... Finding these sequences in purported Ogam inscriptions ... seems to open the door to unbounded subjectivity: I hope that it isn't unduly uncharitable to say that I could produce 'Celtic text' based on these principles for virtually any series of letters (or strokes) which you supplied" (Carey, personal communication).

One other piece of this story remains to be noted. The Wonderful West Virginia editors rely for confirmation of the Fell decipherment on its reported verification by a phenomenon of nature involving the sunrise on the day of the winter solstice.

One of the authors of the Wonderful West Virginia articles, Gallagher (1983) writes that when she first saw the Wyoming petroglyph, the "sun symbol" gave her "reason to wonder if the winter solstice sunrise might be observed from the rock shelter." It was shortly after that that Gallagher got in touch with Fell and arranged to have the petroglyph photographs sent to him. His immediate response was that the markings were in Ogam and that the first sentence on the Wyoming cave wall translated: "At the time of sunrise a ray grazes the notch on the left side on Christmas Day" (Gallagher, 1983; 7-8).

Fell then suggested that the Wonderful West Virginia authors go to the Wyoming rock shelter at dawn on December 22,1982. They did this. Gallagher reports that as the sun rose, "a sunbeam funneled through a three-sided opening in the left side of the rock shelter and struck the sun symbol on the left side of the petroglyph. The rising sun soon bathed the entire panel in light." And again: "The group continued to watch as the solar phenomenon demonstrated physical proof of Fell's decipherment" (Gallagher 1983: 13).

Whatever the full explanation may be of the 18 slashes set in circular form on the edge of the Wyoming rock wall (see Brashler, this volume) the conclusion that the design confirms Fell's "decipherment" is an affront to reason. What was done here is incriminatingly self-evident.

Gallagher's own account leaves little question but that she reported her fairly extraordinary "hunch" to Fell before he ever saw the pictures of the petroglyph. It is equally evident from his report that he then proceeded to find confirmation of that hunch by the most tortured of all his manipulations of the Wyoming rock shelter marking into "Ogam."

A partial illustration is sufficient. The key words in Fell's "confirmatory decipherment" are, in English, "Time of sunrise." He identifies the old Irish equivalent for these words as being "am eriggren." So he needed to find in the Wyoming marking the Ogam script equivalent of the consonants m-r-g-g-r-n. Once he had these he could fill in such vowels as he chose.

Sun Symbol
Portion of Figure L (Fell 1983)
© 1983 WVDNR, used with permission
In what seems almost a taunt at others' gullibility, Fell decided to find those consonants in what he identifies as the "sun symbol" - the 18 slashes extending out in all directions from a central point. He calls this a "rebus," which it isn't by any accepted dictionary definition; but the word adds a suggestion of mystery. The group of slashes is in any event the least Ogam-like on the face of the wall. The lines run in literally all directions, only two of the 18 being perpendicular; and there is no suggestion or possibility of finding here the horizontal stem line which is Ogam's key characteristic.
Rebus 'unscrambled'
Fell's interpretive method 'unscrambles' the portion in the photograph (Fell 1983: Fig. M-1)
© 1983 WVDNR, used with permission

Using Fell's unprincipled technique, this presented no problem. He found his m-r-g-g-r-n by taking his three standard liberties. He made perpendicular lines of those that were cut at all angles. Then he constructed a stem-line that wasn't there, deciding arbitrarily to run it through the middle of the "perpendicular" lines. Finally, he divided these lines, again arbitrarily, into six groups - 1,5,2,2,5 and 3. Presto! This gave him Ogam m-r-g-g-r-n.

There would have been equal justification, or lack of it, for drawing in his stem line below the "perpendicular" lines and dividing them into 4,2,1,3,3,4,1 groupings-which would have given him, in "Ogam," the consonants c-d-h-t-t-c-h. Or he could have added his stem line above the lines he had straightened up; grouping them 4,3,2,1,3,4,1 would have produced s-f-l-b-f-s-b.

The only identifiable reason (and he offers no other) for Fell's arbitrary manipulation of the slashes on the rock wall was that they gave him the consonants he needed for the Ogam version ... adding three vowels of his own choosing ... of the Old Irish phrase that could be translated into "time of sunrise." Doing the same thing with fifteen or twenty markings above and below the "sun symbol," he got his confirmation of Ms. Gallagher's initial speculation. Fell derived "Christmas Day" from three lines, none of them either horizontal or perpendicular, which he "deciphers" as the letter "D" in Ogam - and translated in Old Irish as "Birthday of Christ."

Dr. Fell's work is an affront to the readers of Wonderful West Virginia and the public at large. The fairest verdict on his hoax is offered by Professor William Gillies, of the Department of Celtic Languages at Edinburgh University in Scotland. Reviewing the Wonderful West Virginia account, he writes that Fell's "decipherment" belies all that is known regarding "Ogam script, epigraphy, Celtic language and Christianity . . . . I find the professed content of these 'inscriptions' far-fetched and the linguistic 'reconstruction' an absolute bar to credibility" (Gillies, personal communication).

Professor Gillies concludes, however, "I suspect there may be more valuable things to be said about these carvings than the preposterous constructions you have had to put up with so far" (Gillies, personal communication). It would be unfortunate if the continuing search for whatever may be these petroglyph's explanation were deterred or obstructed by what turns out to be a demeaning fraud.


The authors acknowledge gratefully the assistance received from linguists and archaeologists of whom we inquired about the Fell "decipherments" of the West Virginia petroglyphs. This list includes, with the dates of their replies:

Carey, John, Department of Celtic Language and Literature, Harvard University. 1987 Letter to authors, April 14.

Chathain, Proiseas, Ni, Department of Early Including Medieval Irish, University College, Dublin. 1983 Letter to Joseph J. Snyder, American Committee to Advance Study of Petroglyphs and Pictographs, November 15.

Gillies, William, Department of Celtic, Edinburgh University. 1987 Letter to authors. September 24.

Herrity, Michael, Archaeology Department, University College, Dublin. 1987 Letter to authors. April 28.

Nicolaisen, Wilhelm F.N., Department of English, State University of New York at Binghamton. 1987 Letter to authors, November 13.

O'Dochartaigh, Cathain, Dublin Irish Language Institution, Dublin. 1982 Letter to Joseph J. Snyder, November 15.

Radner, Joan, Department of Literature, American University, 1987 Interview.

References Cited

Cook, Warren L.
1977      Ancient Vermont, Proceedings of Castleton Conference, October 14-15, 1977, Academy of Books of Rutland College.

Daniel, Glyn
1977      Review of America B.C., New York Times Book Review, March 13.

Fell, Barry
1976      America, B.C., Ancient Settlers in the New World, Quadrangle/New York Times Book Co.

1983      Christian Messages in Old Irish Script Deciphered from Rock Carvings in W.Va., Wonderful West Virginia, March.

Fitzhugh, William H. and Ives Goddard
1978      A Statement Concerning America B.C., for the Department of Anthropology, Smithsonian Institution, Man in the Northeast, pp. 166-72.

Gallagher, Ida Jane
1983      Light Dawns on West Virginia History, Wonderful West Virginia, March.

Hyde, Arnout, Jr.
1983, 1986      Editor, articles in Wonderful West Virginia, March, 1983, June, 1986.

CWVA Home Page
"Ogham Index" Introduction Page

© 2003 CWVA

West Virginia Archeologist: Hunter Lesser on cult archaeology
West Virginia Archeologist: Oppenheimer and Wirtz look at Fell's methodology
Wonderful West Virginia:      Robert L. Pyle on pre-Columbian contacts
West Virginia Archeologist: Hunter Lesser looks for pseudoscience, and finds it
Wonderful West Virginia:      Ida Jane Gallagher finds a "solstice alignment"
West Virginia Archeologist: Roger Wise looks for that "solstice alignment" again
Wonderful West Virginia:      Barry Fell deciphers Christian messages
A second opinion:                  A very different translation of Horse Creek
West Virginia Archeologist: Janet Brashler looks for all possible explanations, and tests them