Oxford University Press's
Academic Insights for the Thinking World


For a long time I have been dealing with the words bad, bed, bud, body, bodkin, butt, bottom, and their likes. The readers who have followed the discussion will probably guess from today’s title that now the time of path has come round. Few English words have been explored so often and in such detail as path, but scholars are nowhere near consensus on its origin despite the richness of the available material and the excellence of the research devoted to it. As usual, I’ll give very few references, for the entire list of the works used for this post can be found in my bibliography of English etymology. In the rare cases when someone wants to follow my sources, I send the questioner the text of the required article (many publications in the bibliography are hard to get, while I keep all of them in my office).

The original pathfinder
The original pathfinder

Path was doomed to give language historians trouble because the word is old and begins with p. This is a fatal combination. Since path was recorded in Old English and has cognates in German and Dutch, it could have been expected to go back to Indo-European antiquity. But the Germanic consonant p should, theoretically, correspond to non-Germanic b, just as t corresponds to d (compare English two and Latin duo). For the reason never clarified to everybody’s satisfaction, reconstructed Indo-European does not seem to have had b. Such gaps in the system are also known in living languages. Those of little phonological faith are always reminded that Modern Dutch has b and d but lacks g. Not everybody agrees that ancient Indo-European did entirely without b, but, in any case, it was very rare.

By contrast, p occurred regularly, and many words in Sanskrit, Latin, Greek, and Slavic that begin with p are semantically compatible with path. The conclusion suggests itself that path was borrowed from some other language. The most probable lender is Avestan pad, which means the same as path. (Avestan, or Zend—its older name—is a dead Iranian language related to Sanskrit.) Many etymologists believe that path is indeed a borrowing of Avestan pad. The idea that path goes back to Celtic has few supporters. We needn’t choose between the two sources. It is more important to decide whether path can under no circumstances be native.

First, something should be said about the distribution of Germanic path on the map of the world. As noted, it occurred in Old English. Its Modern German and Dutch cognates are Pfad and pad. Pfad is a bookish word and is almost unknown in southern dialects. Dutch pad corresponds to path and means “path, way.” From it we have pad “robber” (gentlemen of the pad, along with footpads, attacked and robbed people). In the scholarly literature, path is often called a West Germanic word. Yet a similar-sounding word (pada) occurs in some Swedish dialects in both Sweden and Finland. There it means “a piece of low ground” or “a small bay.” For this reason, Hans Jonsson, the author of a detailed book on the Scandinavian names for “body of water,” declared path ~ pad ~ Pfad and Scandinavian pada to be different words (homonyms), though he knew that one of the senses of Old Engl. pæþ (pronounced as path in American English) was “valley.” When one combines all the recorded senses of the West Germanic word, one comes up with “way; valley” and “swamp.” The Swedish forms fit this array quite well. I doubt that Swedish pada can be separated from its West Germanic look-alikes and will now ask: “How could such three meanings coexist in one noun?”

Paderborn on the river Pader.
Paderborn on the river Pader.

Albrecht Greule, another distinguished student of hydronyms, concentrated on the river name Pader, known to many from the name of the town Paderborn in Westphalia. He looked at the root pat- ~ bat- and came to the familiar result that it meant “to swell.” In some older posts I mentioned this root more than once. It was isolated long ago, and numerous words, from Engl. puddle and pudding to pad “frog” and bud and button found themselves in this etymological orphanage. Whether the obscure English noun pod belongs here is anybody’s guess. “Swell” is a good meaning on which to base a river name, but “valley” and “path” look less promising.

I am returning to Wilhelm Oehl, whose ideas figured prominently in the post of 21 October. One of his articles deals with the word for “foot” in the languages of the world. From Bantu to English, sound groups like pita, bode, fat, feta, and their likes, all meaning either “foot” or “tread,” occur with astounding regularity. I’ll mention German Pfote “paw,” from pote; several Romance forms sound alike. Among many others, the non-Germanic congeners of Engl. foot are Latin ped– (as in pedestrian) and Greek pod– (as in podium). Old Icelandic yields more words of the same type (Oehl cited them). It is hard to deny the idea that we have before us a sound-imitating group. Hardly anyone will disagree that Engl. pit-a-pat is an onomatopoeic formation. Some time before Oehl (in the 1920s) Ferdinand Sommer came to the same conclusion, though his material is not so rich. Unfortunately, his article was published posthumously in 1977. Elmar Seebold, the editor of the latest editions of Kluge’s etymological dictionary of German, is aware of Sommer’s work, but he may not have read Oehl. He gives cautious references to Sommer in the entries Pfad and Pfote. Oehl’s results would have spared him the uncertainty.

In the past, I have noted more than once that the concept of homonymy is inapplicable to sound-imitative and sound symbolic words. Those are like imaginary numbers, which cannot be larger or smaller. It appears that all over the world people use the complexes pad ~ pat and bot ~ bud (with various vowels between them) to designate swelling and the same complexes to refer to treading and the foot, the organ of treading. Perhaps the river Pader “swells,” while valleys and swamps “expand.” Unlike those, paths were meant for padding along (compare the verb paddle “to walk through the mud”).

No easy path leads from Avestan to English.
No easy path leads from Avestan to English.

This hypothesis does not make further research into the history of every individual word useless. Oehl’s tendency to offer only a bird’s-eye view of the material has its merits (the general picture becomes clear) and demerits (too many heterogeneous words are lumped together without discussion). German Pfote has unquestionable analogs in Romance, but we still don’t know whether Germanic borrowed the Romance noun or Romance borrowed it from Germanic, or whether both go back to a third language (some substrate), or, finally, whether the two language groups coined the words individually, that is, independently of each other. In my opinion, path and its Germanic cognates were not borrowed from Iranian. As Theodora Bynon pointed out, if the word had come to West Germanic from Avestan, it would have left some traces between England / the Netherlands and Iran, but despite her attempt to find the Celtic source of path, the phonetic difficulties are too serious to be ignored.

So a primitive local formation? A migratory word? Perhaps.

Image credits: (1) The Pathfinder by James Fenimore Cooper (Modern Library Edition) via Pat Iacuzzi Pinterest. (2) Fontes Paderae, seu Paderborn (1671). Monumenta Paderbornensia 2. Ausg. 1672, S. 168 UB Paderborn. Johann Georg Rudolphi, Romeyn de Hooghe. Public domain via Wikimedia Commons. (3) Bodleian Library, MS J2 fol. 175. Public domain via Wikimedia Commons.

Recent Comments

  1. Paul Nance

    Very interesting and thought-provoking. Does the fact that German Pfad and Pfote both begin with pf- suggest borrowing? I seem to remember reading that all German words with initial pf (Pfeffer, Pfarre, Pfund, etc.) are early borrowings from Latin. Or is that an over-generalization?

  2. Jevgēnijs Kaktiņš

    I enjoyed “The calf-path” by Sam Foss. Just wondering whether paving the way, cowpaths (what is the origin for “don’t pave the cowpath?”), building roads and bridges would not ultimately lead us to language architects of Rome. Well, Latvian teļš (calf) and ceļš (road) are not etymologically related.

  3. Patrick Taylor

    Pad is not the Avestan word for “path”. The Avestan word for “path” is attested as paṇtā̊ in the nominative and has an oblique stem paθ-. That aside, it would be better to say that the Germanic word is borrowing from an East Iranian language with which Germanic was in direct or indirect contact in antiquity. Avestan in an Eastern Iranian language, but Iranists generally generally locate Avestan (as the living, spoken language of the community in which the development of the Zoroastrian religion occurred and of the composers of the Old Avestan hymns and liturgical texts) in Afghanistan and adjacent regions. If the Germanic word is of Iranian origin, then it was doubtless borrowed from one of the Scythian languages (also Eastern Iranian) of ancient eastern Europe, not directly from Avestan.

  4. awno ka

    I am not a scholar but the Egyptian deity Ptah seems like a likely common origin?

Comments are closed.