Frequent commenter Paul sent me this quote from Robert Macfarlane’s The Old Ways: A Journey on Foot, a book he recommends at least as heartily as Rob Nixon in the Times (“lively, luminous… gorgeous”):

It’s true that once you being to notice them, you see that the landscape is still webbed with paths and footways—shadowing the modern-day road network, or meeting it at a slant or perpendicular.
Pilgrim paths, green roads, drove roads, corpse roads, trods, leys, dykes, drongs, sarns, snickets—say the names of paths out loud and at speed and they become a poem or rite—holloways, bostles, shutes, driftways, lichways, ridings, halterpaths, cartways, carneys, causeways, herepaths.

It’s hard to resist an author with that kind of feel for words.


  1. And here’s the delightful etymology of path from the OED, which shows once again the thickets in which would-be etymologists, even those equipped with the most modern lamps, guns, and cameras, can wander:
    Etymology: Cognate with Old Frisian path (West Frisian paad, East Frisian (Saterland) pad, East Frisian (Wangeroog) path), Middle Dutch pad, pat (Dutch pad), Middle Low German pāt, pat, Old High German phad, pfad (Middle High German phat, German Pfad); further etymology uncertain, perhaps borrowed early < an Iranian language (see note below).
    The word is apparently restricted to West Germanic; there is no evidence in Gothic or the early Scandinavian languages (but perhaps compare Finnish pade valley, probably < a Germanic language). The forms show that the word must have been in West Germanic before the Christian era.
    The form of the consonants is problematic. While the final fricative suggests the regular operation of the First Germanic Consonant Shift (Grimm’s Law), the origin of the initial p- is debated: according to Grimm’s Law, an underlying Indo-European p- should have shifted to f- ; alternatively, Germanic p- could derive from Indo-European *b- , the existence of which is uncertain.
    The most widely accepted theory sees the word as a borrowing from Iranian, in which Indo-European p- is preserved, and there is alternation between forms with -t- and forms with -θ- ; compare Avestan pantā (nominative), paθō (genitive) way, Old Persian pathi– , ultimately < the same Indo-European base as find v. (compare found v.1). This explanation does however pose historical problems, given the limited distribution of the Germanic word.
    An alternative suggestion assumes a borrowing from an unattested Gaulish term (< the same Celtic root as Old Welsh, Welsh pant valley, of unknown origin; compare sense 2a). While this model can account for the consonants, the vowel quantity is unexplained (a long vowel would be expected).
    H. Kuhn (Zeitschr. Mundartforschung (1961) 28 4, 14) lists the word with a small number of West Germanic terms with unshifted initial p-, which he regards as deriving from a pre-Germanic Indo-European substratum; he regards path as being ultimately < the same Indo-European root as foot n.
    In Old English the stem vowel typically alternates between æ and a, with the latter occurring before endings containing a back vowel, and the former occurring elsewhere (see A. Campbell Old Eng. Gram. (1959) §157); however, there is much analogical levelling. Old English forms with o as stem vowel are attested only when the word is the second element of compounds, and result from low stress (see A. Campbell Old Eng. Gram. (1959) §335). The evidence sometimes adduced for an Old English feminine by-form paþu is late and doubtful.

  2. He’s missed out vennel, close, wynd and alley.

  3. We did vennels, gennels, and snickets here.

  4. LH We did vennels, gennels, and snickets here.
    I read that entire thread and found gennel/ginnel, snicket and alley, along with a few others wuch as whin and twitten/twitting, but not vennel.
    I believe that vennel must be from the old French word la venelle. I know the word from reading it, in Balzac, for instance, where the context is always pejorative. Without looking this word up in a dictionary I visualize a very narrow and dark medieval street with a shallow gutter in the middle, found in a slummy area inhabited not only by the poorest of the poor but also by petty crooks, etc.
    Now (after consultation) the TLFI just says “a small narrow street”, but the example given mentions that this street leads into an area of (vegetable) gardens, outside of the town. I think this is one of the meanings given for snicket in the earlier thread, a passage from an urban area into an open one.

  5. I looked up lichway to find that it’s the path along which a coffin is carried to the grave. Knowing lychgate, I googled that, and found (as I knew) in wikipedia that it is a covered gate leading to a churchyard. Lo! and behole! there is a photograph showing a graveyard on the other side of the gate.
    Lychgate can also be spelled lichgate.
    Further googling, and I found that is used in fantasy fiction to refer an undead person, but I can’t find any other use.

  6. I read that entire thread and found gennel/ginnel, snicket and alley, along with a few others wuch as whin and twitten/twitting, but not vennel.
    It’s in the quoted OED entry for snicket: “1947 I. Brown Say the Word 65 We have vennels, gunnels, and snickets in our northern towns.” We didn’t actually discuss it, but it was there!

  7. And yes, the OED (as of 1916) says it’s from “Old French venele, venelle, vanelle (modern French venelle) < Romance type *vēnella (medieval Latin venella), diminutive of Latin vēna vein.”

  8. It can mean both “A narrow lane, passage, or thoroughfare in a town or city; an alley or wynd. Chiefly Sc.” (1879 N. & Q. 5th Ser. XI. 137/1 “In the town of Strabane, Ireland, there are a number of narrow passages, called ‘vennels’, from the main street to the river shore”) and “north. An open drain or gutter; a sewer” (1881 J. Sargisson Joe Scoap’s Jurneh 93 “Carry’t t’ watter off beaath ways inteh t’ vennels”).

  9. Jeffry House says

    lichgate reminds of Norwegian “lik” meaning corpse, and “gate” meaning street.
    I can’t find “likgate” anywhere on line, though.
    The older word would have been “lig”, as it still is in Denmark. That terminal “g” often transmutes into a “ch” sound, as in our example.

  10. The vennel I grew up with had the full title “Friars’ Vennel”.

  11. Norwegian . . . “gate” meaning street.
    English has gate meaning entryway and also lane; German has Gasse, narrow street or alley. These and more, including go/going/gone, from PIE ghe-.

  12. Thanks, JH. So the word comes out of the old Danelaw, I presume.

  13. Iakon: The native word for ‘gate’ survives as yate in certain compound proper names and in the surname Yates, Yeats. As for lic > lich, the modern adverbial ending -ly is from lic + -e, the OE adverb ending (now lost or reduced to zero in “flat adverbs”). So English slowly is literally ‘with slow body’, whereas Spanish lentamente is literally ‘with slow mind’ < L. ablative of mens.

  14. Re rareness of lych/lich, I just want to mention the Lyke-Wake Dirge, with always gives me goosebumps:
    This ae night, this ae night,
    Every night and all,
    Fire and fleet and candle-light,
    And Christ receive thy saul.

  15. Jeffry House says

    I had no idea that the modern ly adverbial ending “ly” comes from lic! That might also explain why in Norwegian it is “lig”, as in “hygge-lig”, pleasant body, Etc.
    In the last few hundred years, though, “lig” has meant corpse, not body, without the dual meaning that English “body” exhibits.
    In Norway, a living body is a “kropp”, from corpus, but has nothing to do with “corpse”. This reminds me of the incident with the Norwegian Blue. But I digress.

  16. It is a goosebumpy song. Here‘s a North Yorkshire variant (e.g., “Christ tak up thy saul”).

  17. @John Cowan: Are “flat adverbs” the result of a process of loss? Sweet/Davis says: “-e is the regular adverb termination […] From the frequent combination of this ending with the adjectival -lic, the suffic -lice is often used to form adverbs […].”

  18. RC: Do you mean the loss of final -e which is explainable through the loss of this sound in final position, regardless of the type of word, or the loss of -lic(e)? The latter would not be credible, since -lic(e) ended up as -ly.

  19. Flat adverbs are the result of the general loss of -e, though I have not yet found a specific example that has not changed its semantics from Old to Modern English.

  20. @marie-lucie: I was evidently too elliptical, as I often seem to be. What I meant was to question what I took to be John Cowan’s statement that “flat adverbs” arose via loss of –lice, though when I reread his comment in the light of hs later one, I’m not sure that’s what he meant.

  21. RC: (JC) the modern adverbial ending -ly is from lic + -e, the OE adverb ending (now lost or reduced to zero in “flat adverbs”)
    I found this statement quite clear that “the OE adverb ending” referred to “-e”, but I already knew this ending from having studied some of the history of English. I can see where this statement could also be interpreted to mean that the ending was “-lice”. It must have been the loss of final “-e”, causing confusion between adjective and adverb, which contributed to the generalization of “-lic(e)” and later “-ly”.
    JC: Flat adverbs are the result of the general loss of -e, though I have not yet found a specific example that has not changed its semantics from Old to Modern English.
    What about loud from OE lhude (as in sumer is i-cumen in, lhude sing, cuccu! ‘Summer has come, sing loud, cuckoo!’). This word could be an exception though: I think that the adverb loud coexists with loudly, with a slightly different meaning and usage: Speak louder! not “loudlier”.
    There is also long in long ago as well as long-lost, long-forgotten, long-suffering etc, where long cannot be considered an adjective.
    I think that the changes of meaning for some of the flat adverbs come from their being used as modifiers, usually intensifiers, in specific adjective phrases, and picking up some of the meanings of the adjectives.

  22. Yes, thank you for the example, m-l. We have loud adj. < hlud, without the ending, and loud adv. < hlude with it. Another example is læt/læte, but this meant not ‘late’ but ‘slow’, so there has been a semantic shift. In other cases, the -e seems not to have been present in OE as we know it.
    Long in the compounds you mentioned is still adjectival, at least diachronically: long-lost means ‘lost for a long time’, just as long-bearded, say, means ‘having a long beard’. The magic is in the participle.

  23. JC, I have to disagree with you about the status of long in those examples: if it means ‘for a long time’, used with a verbal form (a participle, as also in longstanding) it is an adverb; if it just means ‘opposite of short’, used with a noun derivative, it is an adjective. It is the word it modifies which (synchronically) determines the status of a word like long. I am not sure why you think that it was different historically, when adverbs ended in -e (which was pronounced, as schwa).

  24. A letter in this morning’s Telegraph says that the York expression “snickelway” would be “twitten” in Sussex.

  25. m-l: I had a brain fart and wrote diachronically when I meant synchronically, so that was a red herring.
    But it is simply not true that every word that modifies an adjective is necessarily an adverb. Adjective-participle compounds like red-bearded, wiggly-fingered, cold-hearted, ultraviolet-reflecting are a regular part of English syntax, and they don’t even require the existence of appropriate verbs. Beard and finger exist as verbs, but they are semantically wrong (they mean ‘defy’ and ‘touch’ or ‘identify’ respectively); as for heart, it is not a verb in current use at all (except as a verbal representation of ♥ in “I ♥ NP” sentences). The first adjective can also be replaced by a noun, as in iron-girded, lead-footed, lake-bordered.
    Both patterns are productive: there is no OED, ODO, AHD, or entry for iron-girded, but plenty of Google hits, and I just made up lake-bordered myself on the analogy of river-bordered. Gerard Manley Hopkins in his poem “Duns Scotus’s Oxford” described the city as “cuckoo-echoing, bell-swarmèd, lark-charmèd, rook-racked, river-rounded”, which is extreme but definitely not ungrammatical.
    Similarly, the existence of simple noun-adjective compounds like grass-green and stone-hard do not entail that grass and stone can be construed as adverbs.

  26. Thanks, JC! That was very illuminating.

Speak Your Mind