Thanks to a comment from Aldiboronti in this Wordorigins thread, I learned an interesting fact: the word truce is essentially just the plural of true. As the OED puts it:

[ME. trewe and triewe, mostly in pl. form trewes and triewes:—OE. tréow n. masc. (fem. pl. tréwa), ‘truth or fidelity to a promise, good faith, assurance of faith or truth, promise, engagement, covenant, league’, = OEFris. tríuwe, OWFris. and MDu. trouwe (Du. trouw), OS. treuwa, tríuwa, OHG. tríuwa (MHG. triuwe, Ger. treue):—WGer. *trewwa, Goth. triggwa ‘covenant’ (whence late L. and Romanic tregua, treuga, F. trève); also, in ablaut form, OE. trúwa n. masc. and pl. –an; = ON. trúa, trú, Norw. trū, Sw. trōa: see TRUE a. Already in OE. the pl. tréwa was often used in the sense of the sing.; this became still more frequent with the ME. pl. trewes, triues, triwes, trues, and finally this, as trews, trewse, truse, truce, became the received sing. (app. in reference to the pledges or engagements given by both parties), with a new pl. truses, truces, when required. Cf. cherries, pease. See also trève from French, and the rare treuges after MLat. treugas.]

But what I want to know is: if it’s originally a plural, why doesn’t it have a voiced final consonant? In other words, why isn’t it trewes or trues?


  1. The entry for die provides a pretty good clue:

    As in pence, the plural s retains its original breath sound, probably because these words were not felt as ordinary plurals, but as collective words; cf. the orig. plural truce, where the collective sense has now passed into a singular. This pronunciation is indicated in later spelling by –ce: cf. the umlaut plurals lice, mice, the inflexional forms hence, once, twice, since, and the words ice, nice, advice, device, defence, in all which –ce represents a phonetic and original –s.

    So it sounds like truce was reanalyzed as a singular form before final /s/ voicing occurred, and the voicing only affected inflexional s‘s.

  2. Exactly. Another example is deuce from (Old or Middle) French deux ‘two’ (currently pronounced with a final [z] between vowels but must have been [s] earlier in all positions). (The x was never pronounced [ks] in this and other words in eux).

  3. Yes, it’s for the same reason that ‘peace’ is simply the plural of ‘pea’–wince quantities of that plant were commonly used as treaty-offerings “in medieval times”.

  4. As Birdseye used to say, “Peas is our profession”.

  5. How far back ago were -s plurals actually pronounced as /s/ and not /z/?

  6. Dice is another one – Alea iacta est: “The die is cast”.

  7. How many modern people in practice regard dice as a plural word? It’s a long time since I played dice games on a regular basis, but I’m pretty sure that when I did everyone treated dice as singular (e.g. “the dice has fallen on the floor”) and if we needed to refer to more than one of them they were dices. The SOED recognizes this usage, calling it colloquial but also attributing it to Late Middle English, so it’s been around a very long time. My impression is that ordinary people only talk about a die in the set phrase “the die is cast”.

  8. I have a different impression about die/dice, and it’s back up by some surely-flawed Google research: “the dice have” gets three times as many hits as “the dice has,” and “the dice were” gets six times as many as “the dice was.” (The pattern was reversed when I searched “the dice is” vs “the dice are,” but most of the hits didn’t seem relevant.)

  9. You never see dice except in pairs. Even out in the boonies where I grew up, we knew the singular of “dice” was “die”–even if we had never seen just one.

  10. @Athel Cornish-Bowden: most dictionaries regard dice as singular now, and don’t even qualify that with “colloquial”.
    The one-volume Oxford Dictionary of English and the Oxford Dictionary of Phrase and Fable have “a small cube …” as does Chambers. Collins defines dice in the plural before saying it also functions as a singular.
    Even the OED, which can take decades to catch up with changes, has “The form ‘dice’ (used as pl and sing) is of much more frequent occurrence in gaming and related senses than the singular ‘die'”

  11. A.J.P. Crown says

    It’s funny that we talk about one dice and two dice (does anyone say dices?), because we’re now also moving in the opposite direction; with computers we have one mouse and two mouses.

  12. In my experience, people who grew up in the role-playing game culture (for example, Dungeons and Dragons) are very used to using “die” in the singular and “dice” in the plural. This is possibly because the game books refer consistently to things such as “a twenty-sided die”. There’s possibly a nerd factor here, too.

  13. @Conrad: and hence the well-known slogan “Visualise whirled peas”.

  14. Yes, one of my favourite dishes as it happens; Ferran Adriá has made a specialty of it.

  15. marie-lucie says

    with computers we have one mouse and two mouses.
    That’s what I thought, but at Staples (at least at my local branch, in Canada) they sell mice.

  16. A.J.P. Crown says

    In A Day At El Bulli most of the recipes require many out-of-the ordinary kitchen appliances, such as a Pacojet, freeze-dryer, liquid nitrogen tank, candyfloss machine, car exhaust and Perspex molds…Criticism has split top Spanish chefs into pro and anti Adrià camps. Unusual dishes that have been criticized include tobacco-flavored blackberry crushed ice and Kellogg’s paella (Rice Krispies, shrimp heads and vanilla-flavored mashed potatoes).
    I like rice krispies ok, but shrimp heads?

  17. Crown, A.J.P. says

    All right, I admit I added ‘car exhaust’, but the rest is from Wiki.

  18. most dictionaries regard dice as singular now
    Most U.K. dictionaries, maybe. Not American ones.

  19. Let’s not forget that “tool and die” manufacturing doesn’t necessarily have to do with gaming dice.

  20. And of course rice is the plural of rye.

  21. “How far back ago were -s plurals actually pronounced as /s/ and not /z/?”
    Final –s voicing must have taken place after truce was reanalyzed as a singular, and that had taken place by the early 1500s, according to the citations in the OED. I’m having a hard time pinning it down more exactly at the moment.

  22. Crown, A.J.P. says

    And spice is the plural of spy?

  23. Then slice must be the plural of sly.
    And if thrice is the plural of three, then nice must be the plural of knee.

  24. In the videos linked at this recent Metafilter post, a man who manufactures dice and seems to know a lot about them uses “dice” as both singular and plural.

  25. As I’m sure nobody here needs reminding, “peas” (or rather “pease”) was originally the singular, the plural being “peasen”, but “peas” was reanalysed as a plural and the singular became “pea”, “pease” surviving only in the name of the dish, pease pudding (mmmm … pease pudding …)
    And as I’ve commented elsewhere, since decimalisation it’s not uncommon to hear in the UK someone say “one pence”.

  26. Not long after decimalisation, in 1971, I heard that the Ladies’ at Harrods in London had a hand-written sign saying “Please use 1 new pence pieces” (in the days when the cubicles were coin-operated. I don’t know if they are still. Hence the expression “spend a penny”.) A pedantic customer pointed out to the attendant that “pence” is plural, and “penny” the singular. The response was “Well, it’s wurf two an’ ‘arf, yer know!” (For youngsters among us, and those unfamiliar with pre-decimal British currency, there used to be 240 pence to the pound, so a new penny was equivalent to 2.4 old ones.)

  27. Arthur Crown says

    there used to be 240 pence to the pound, so a new penny was equivalent to 2.4 old ones
    Thanks for that, I’d never thought of it that way before.

  28. marie-lucie says

    “peas” (or rather “pease”) was originally the singular, the plural being “peasen”
    Was peas really the singular (one peas), or could it have been the mass noun, as in pease porridge, with peasen used for a small, countable amount of individual peas? (there isn’t enough for a decent bowl of porridge, only five or six peasen left in the bag, that sort of thing)? This reminds me of oats, a mass noun, from which a singular oat is sometimes back-formed.

  29. Marie-Lucie:
    Indeed pease was a singular countable to start with, if OED is to be believed:
    1. Singular.
    […] 13.. K. Alis. 5959 A pese nys worth thi riche slaunder. 1362 Langl. P. Pl. A. vii. 155 A wastour countede pers at a peose [1377 B. vi. 171 pees] and his plouh boþe. c1380 Sir Ferumb. 5847 By Mahoun y nolde Šyue a pyse, for cryst ne al ys myŠte. 1390, c1400 Pese [see B. 2]. 1483 Cath. Angl. 273/1 A Peise, pisa. 1530 Palsgr. 158 Vne poyx, a pees. 1580, etc. Pease [see B. 2]. 1614 Raleigh Hist. World i. iv. §2 Of the bigness of a great Peaze.
    (Unless OED’s revision of the P words has since changed this story. I haven’t checked the online version.)
    From Latin pisum, and earlier Greek.

  30. marie-lucie says

    Thank you, Noetica. The word is unlikely to have come directly from Latin pisum or it might now be pise, it must have gone through an earlier version of French pois which must have been peis originally, itself from the Latin word.

  31. Marie-Lucie:
    Strangely, no. OED has it coming directly from Latin, before any plausible French influence. It does not mention French in the etymology. There was already very early slippage between singular and plural forms and meanings. Anyway, pise was indeed an early spelling. I omitted a source with pise, above:
    c1050 Cotton Cleop. Gloss. in Wr.-Wülcker 432/25 Lenticula, pise.
    And OED gives that spelling first in its etymology:
    [OE. pise (piose, pyse) wk. fem., pl. pisan, a. L. pisa (pl. -æ), late collateral form (4th c. in Palladius) of pisum, pl. pisa, a….]

  32. Ah, when I say a singular countable to start with, I mean without dependency on an earlier plural acceptation in English. OED has the word as plural coevally with the word as singular.
    As for Latin and French, I do not claim anything other than this: OED does not implicate French in the etymology. As you can see from what I quote, there a Latin form pisa, intermediate between pisum and OE pise.

  33. Noetica, thank you for verifying the information. I don’t have a dictionary other than the Petit Robert, which is not exactly informative about English, and I am too lazy to take the trouble of looking up the OED!.
    I am still unconvinced though: OE pise, piose, pyse should have ended up still written pise and pronounced like wise. The current vowel is more likely to have been influenced by Old French peis, at a time when the final s was pronounced (ea at the time was a lower vowel than now). The Middle English sources you cited earlier have pese, pees, peis, etc, all of which are closer to the OF word. Perhaps the closeness of the two pronunciations caused a change in English?

  34. The etymology from the OED draft revision dated September of this year:
    [< post-classical Latin pisa (feminine; plural pisae) variety of pea (4th cent.), variant of classical Latin pisum (neuter; plural pisa), either < ancient Greek πίσος, also πίσον pulse, pease (apparently a loanword), or representing a parallel loan from another language; subsequently reinforced by Anglo-Norman peis, pais, peas, peiz, pois, poais, peses, poyes and Old French pois, (Normandy) peis, Middle French pois (c1160 as peis; French pois).
      By late Middle English the plural form was often identical with the singular (compare singular forms above and β forms of the plural); subsequently, the final voiced sibilant of the plural form came to be analysed as the mark of the plural ending and a new singular PEA n.2 was inferred, while the earlier forms mostly fell into disuse in standard English (but compare PEASEMEAL n., PEASE PUDDING n.). After the mid 17th cent. the plural (and collective) form peas (see β forms of the plural) becomes indistinguishable from the plural of PEA n.2 The δ forms of the plural [pesons] represent a morphological double plural.
      The modern form represents, on the one hand, a reflex (with Middle English open syllable lengthening) of early Middle English pesan, etc. (with short e, perhaps preserved in Middle English pesene, pessen; < an unattested form in Old English (Anglian) with –eo-: compare Middle English (west midland) peose, peosen, etc.) and, on the other hand, the Anglo-Norman form (with the usual development of the Anglo-Norman diphthong to open ē: compare e.g. PEACE n.). The earlier form of the Anglo-Norman diphthong is reflected in Middle English peise, peysse. Middle English poysyn perhaps reflects Central French pois. The Scots (and probably the Irish) forms apparently represent forms with northern Middle English short i (from Old English (Anglian) piose, piosan) in pis, pizz, etc., and probably also, with open syllable lengthening, in peese, peice etc. Middle English pisen, etc., however, sometimes go back to forms without velar umlaut (compare Old English (West Saxon) pisan).]

  35. Thanks LH. Both Marie-Lucie and I were too lazy to check the latest from OED, and it does turn out to be relevant. Marie-Lucie remains right that French is somehow involved; I was right that OED had not earlier mentioned French at all, and even now brings it in secondarily, beginning with “subsequently reinforced by Anglo-Norman peis,…”. The story remains convoluted. I note that OED now gives earlier first citations. For its illustrations of form for both the singular and the plural, the unrevised entry had the same earliest source:
    c725 Corpus Gloss. (O.E.T.) 1208 Lenticula, piose.
    [singular form]
    c725 Corpus Gloss. (O.E.T.) 1586 Pisum, piosan.
    [plural form]
    These are not in the revised entry, which nonetheless amounts to a significant expansion and improvement. Good to see ”mass noun” used. I wish all dictionaries would efficiently mark the salient distinction between these and countables.
    My main SOED is a CD-ROM from 1997. Does anyone know how the latest SOED handles ”pease”? Mine does not mention French.

  36. marie-lucie says

    Noetica, so we are both right.
    Influence of similar words in a situation of bilingualism (as occurred in the years of French-speaking dominance in medieval Englans) often leads to hybrid forms. An example in French is the word haut ‘high’ which is from the Latin adjective altus, alta, altum blended with the h-initial Frankish (Germanic) equivalent at the time of Frank dominance around 5-600 AD, hence the “aspirated” h at the beginning of the French word (but not of the Spanish or Italian equivalents which preserve the Latin pattern). In the case of peas the English and French forms were sufficiently similar that people were no longer sure which was which.

  37. Sounds right to me, Marie-Lucie. And I trust your authority!
    O, and I should have written “I note that OED now gives later first citations”. Curious, that – in a revised and expanded entry.

  38. David Marjanović says

    the h-initial Frankish (Germanic) equivalent

    Modern German hoch; if the vowel has stayed the same over the last 1500 years, it fits the French one. That said, is the French one that old? When did a in front of l begin to get rounded?

  39. marie-lucie says

    the h-initial Frankish (Germanic) equivalent
    Modern German hoch; if the vowel has stayed the same over the last 1500 years, it fits the French one. That said, is the French one that old? When did a in front of l begin to get rounded?
    The Latin vowel did not change before l, but the sequence al changed to au before a consonant, then later to [o] (this is also why there is the singular/plural alternation as in cheval/chevaux). I am not sure at what stage the h was added, but this addition has to be due to Germanic influence as nothing else could justify it (all the French words with “aspirate h” are of Germanic origin, and the h is still pronounced in some dialects).
    Frankish domination started around the 6th century but went on for several centuries, for instance we share the Frankish emperor Charlemagne (= Karl der Grosse) with Germany, but the Frankish population eventually merged with the early French-speaking population and the Frankish language became extinct in France itself. This is similar to what happened in England after the Norman conquest, where the general population continued to speak English but the language was greatly influenced by the speech of the conquerors.

  40. John Cowan says

    In fact there are about 20,000 speakers of Frankish left in Frankrijk, mostly in the arrondissements of Lille, Douai and Dunkirk; their home language is West-Vlaams and their Dachsprache is southern Standard Dutch (as well as French, of course). West-Vlaams has markedly different vocalism from Dutch; indeed, almost every vowel has developed differently.

  41. David Marjanović says

    different vocalism

    Also, it’s h-dropping, a feature otherwise found in Germanic only in England, Wendland and the elven dale.

    (There is an English article, but I’m linking to the German one for the map.)

Speak Your Mind