The evolution of the Latin consonant cluster "ct"

Discussion in 'Etymology, History of languages, and Linguistics (EHL)' started by Nino83, Mar 19, 2014.

  1. Nino83 Senior Member

    Hi everyone.
    I was wondering how the Latin consonant cluster ct developed in Romance languages.

    In Italian the result is often the same: ct --> tt while in Western Romance languages there are more result (ct, t, it/ch).

    A few examples (in this order: Italian, French, Spanish, Portuguese)

    effetto, effet, efecto, efeito
    rispetto, respect, respecto, respeito
    perfetto, parfait, perfecto, perfeito
    corretto, correct, correcto, correcto (correto)
    fatto, fait, hecho, feito
    diritto, droit, derecho, direito

    The question is this: is there any rule that make us predict whether there is ct/t or it/ch?

    Thanks to everyone
  2. CapnPrep Senior Member

    The k in clusters like kt vocalized to i/y in French, Spanish, and Portuguese (and triggered palatalization in Spanish). You can see this development in the popular outcomes of factum and directum in your list. For further examples, look at basic vocabulary like noctem, octo, *lactem. The words in your list that retain kt are learned reborrowings.
  3. Nino83 Senior Member

    Thank you!

    Another question: why we have for the same word two different outcomes? For example, perfecto in Spanish but parfait/perfeito in French and Portuguese or respect/respecto in French and Spanish and respeito in Portuguese?

    Were these words reborrowed in a different time? This is not a very persuasive conclusion.

    Why are there these differences?

    In Italian all reborrowings were adapted to the new pronunciation (i.e tt), also in writing.
    Last edited: Mar 20, 2014
  4. CapnPrep Senior Member

    The dividing line between popular and learned vocabulary is not always clear, and each word has its own story: when it entered the language, how often it was used, by whom, and in what situations, and how the word was influenced by the other words in the language. This can easily lead to distinct outcomes for similar-looking words in the same language. French parfait, for example, is still linked to the verb faire, whereas respect is not deverbal, and instead has links with a family of words like aspect, suspect, perspective, inspection, spectacle. But note also the doublet form répit (cf. dépit < despectum).

    Every language has its own history, too, so there is no reason to expect French, Spanish, and Portuguese always to go in the same direction.
    Yes, Italian is generally very boring in this respect. :p However, you can find kt in more recent loans (not necessarily from Latin) and in specialized vocabulary: cactus, factotum, detective, directory, -ectomia, etc.
  5. Nino83 Senior Member

    Thank you CapnPrep.

    So English kept the original ct while Italian changed it to tt.
    It's a pity that French, Spanish and Portuguese hadn't regularized this consonant cluster, because one has to search for every single word in the dictionary.

    It doesn't matter :)

  6. Quiviscumque

    Quiviscumque Moderator

    Ciudad del paraíso
    Dear Nino83, do you mean that we Spanish-speakers, even if we say [eʧo], should write "facto" just in order to alleviate your workload :)?
    Now seriously, the typical Spanish speaker is not aware of the fact that "estrecho" and "estricto" are the same word (if you want to say so). Nor should he, I think.
  7. Nino83 Senior Member

    No, I wasn't so serious :)
    I've just said that French, Spanish and Portuguese writing would be easier if this cluster were more regular, but if it's not like that, it's not a problem.

    The Spanish shouldn't write "facto". I was asking why they didn't regularise all these words, for example, saying efecho, respecho, perfecho (just like the Portuguese say efeito, respeito, perfeito), correcho, instead of mantaining the Latin pronunciation.
    Last edited: Mar 20, 2014
  8. Angelo di fuoco Senior Member

    Russian & German (GER) bilingual
    I'd like to throw in words like auto/acto and cautivo.
  9. berndf Moderator

    German (Germany)
    No it didn't "keep" it. Words like perfect (Middle English parfit) are early Modern English re-Latinizations.
  10. CapnPrep Senior Member

    Could you also throw in some commentary about why you are throwing in these words? Particularly auto, since cautivo is slightly off-topic (< Latin pt, not kt).
  11. francisgranada Senior Member

    The same can be said e.g. in case of the Italian prefixes re-/ri-/ra- (though ra- is a bit different as it comes from re+ad). We have recuperare, responsabile, reazione, rappresentare, rispondere, ritornare, ripetizione ... while in Spanish it's always re-.
    Because the cluster ct does not "spontaneousely" or automatically become ch in Spanish. To obtain efecho from efecto all the phonetical process explained by CapnPrep (#2) should be repeated somehow which already does not work.
  12. Nino83 Senior Member

    So was it more likely that ct becomes it in Portuguese than ch in Spanish? Why?
  13. francisgranada Senior Member

    Maybe in a certain period in the past yes (as from ct to it only one "step" was needed), but today I think no more. There are learned words also in Portuguese where ct has not become it, e.g. reto (<recto), efetivo, afeto, estrito (but also estreito), etc ... All in all, generally I think the Portuguese rather maintains more words like perfeito, while in Spanish they were lost/replaced. By the way perfecho exists in the Asturian language.

    P.S. I can imagine (but I don't know) that some words in Spanish never had the variant with ch, but they were replaced by the Latin variant earlier (i.e. when the it was not yet palatalized to ch) or simply, they never existed in the spoken language.
    Last edited: Mar 21, 2014
  14. Youngfun

    Youngfun Senior Member

    Bắc Kinh
    Wu Chinese & Italian
    The most common mistake for me in Portuguese is that I used to never remember corre(c)to and write correito instead, which coincidentally is the pronunciation in many Brazilian accents.
    So instead of a foreigner, I could pass for an ignorant Brazilian. :D
  15. francisgranada Senior Member

    Is this valid also for other words with original ct?
  16. Nino83 Senior Member

    Right, I didn't think about this step.

    The Cariocas tend to pronounce vowels before final as diphthongs (for example [ũma veiʃ] for uma vez), so it could be a dialectal feature.
  17. Youngfun

    Youngfun Senior Member

    Bắc Kinh
    Wu Chinese & Italian
    Now that I think about it, I think I got confused. Paulistas only add "i" before s (três [treis]) or nasal (viagem [vjaʒeĩ]).
    Not sure if they add it in "correto". So let's wait for a Brazilian to confirm.
    Last edited: Mar 21, 2014
  18. Cossue

    Cossue Member

    Galician & Spanish
    Incidentally, and as a speaker of Galician, a sister language of Portuguese, I can confirm that many clusters ([k.t], [b.s], [p.t], [k.k]...) present in learned words are difficult for many popular speakers, who tend to pronounce them either as a single stop consonant, or as a closed vowel or semivowel and a consonant, so regularizing them to the inherited lexicon. An anecdote: I work in an employment office in western Galicia (in Barbanza), which means that many people come by me in a daily basis asking about the "subsidio" (dole, unemployment subsidy). Now, while most people speaking Castilian Spanish would pronounce it [sub'si.dio], many Galician speakers would say [su:si.di.o] and some even [sui'si.di.o]... which sound exactly as a local would pronounce "suicidio" 'suicide' :eek: (which anyway is another learned word who no popular speaker would use, using "matou-se" "he/she killed him/herself" instead). And although I don't longer hear it, [dow'tor] was in the past the common pronunciation of "doctor".
  19. Youngfun

    Youngfun Senior Member

    Bắc Kinh
    Wu Chinese & Italian
    In Brazil subsídio would be pronounced [subi'sid(ʒ)iu].

    Which is the standard in Portuguese: doutor.
  20. Angelo di fuoco Senior Member

    Russian & German (GER) bilingual
    It's a similar phenomenon: the vocalisation of the first consonant in a consonant cluster, although with a somewhar different outcome. While preceding e seems to convert ct in it- in -ch, a seems to favorise -uCt, be that consonant c (auto), p (cautivo) or (at least, in writing) b, like in ausencia.
  21. bearded

    bearded Senior Member

    To the languages mentioned, I think that also Romanian could be added:
    (usually) ct > pt
    Example: nocte(m) > noapte.
    The reason for this evolution is not very clear, it seems.
  22. CapnPrep Senior Member

    But factum has an a… I think examples like auto and pauta must be non-Castilian dialectal forms.

    The u in cautivo, ausencia is probably linked to the labial articulation of p and b, since you also get it after e, e.g. debta > deuda.
  23. danielstan Senior Member

    Romanian - Romania
    Try to find "reason" in the evolution of a language is just to distinguish possible situations for 1 change like the above:
    1) internal evolution of the language (the "dictatorship" of the speakers who could modify the language in any aspect, unconsciously)
    2) the influence of another language, influence which is received by bilingual speakers

    For the case of Latin group ct - it is remarkable this group was not stable (unchanged) in any Romance language.

    I will not solve the above Romanian case (I believe it is an internal evolution of the language).
    Romanian phonetic rule:
    Latin ct > Romanian pt

    Rule is satisfied by all cases of applicability.
    lat. octo > rom. opt
    lat. lucta > rom. lupta
    lat. directus > rom. drept
    lat. accusative noctem > rom. noapte
    lat. acc. lactem > rom. lapte
    lat. factum > rom. fapt
    I don't know exceptions.
    Albanian(*) phonetic rule:
    Latin ct > Albanian ft
    Rule is satisfied by almost all cases of applicability.
    lat. lucta > alb. luftë
    lat. *trocta > alb. troftë
    lat. cotoneum > alb. ftua
    Exceptions, dependant of the previous syllable:
    lat. directus > alb. dreitë
    lat. tractare > alb. traitoj

    (*) Although Albanian is not a Romance language, it has suffered a great influence from Balkanic Vulgar Latin.
    Some linguists consider Albanian as a half-Romanised language.
  24. Delvo Senior Member

    American English
    ... as can be seen from other examples which were adopted earlier in which "ct" became "t": joint (junct-), point (punct-), saint (sanct-).
  25. Nino83 Senior Member

    actus > act, while in Italian it is atto.

    It seems that Latin learned words in Italian were adapted to the common pronunciation while in French, Spanish, Catalan and Romanian the Classical Latin pronunciation was retained, against the common pronunciation.
    In Portuguese these words were written according to the Classical Latin spelling but the cluster was reduced to a single /t/ in most words and the last acordo ortográfico follows the common pronunciation, see ato and fato (insteid of acto and facto).

    This seems very strange to me, it is like French and Spanish didn't find their pronunciation appropriate, like they felt ashamed of their popular pronunciation.
    Why one would retain a pronunciation that is 1500 years old?
    Last edited: Sep 25, 2015
  26. CapnPrep Senior Member

    But how can you tell that those words are learned in Italian? What would have been the common pronunciation/popular development of actum or factum? As I mentioned in an earlier message (#4 above), Italian does have unassimilated kt clusters in some types of words.
  27. Nino83 Senior Member

    Atto and effettivo are not "popular" words, but they were reintroducted with the popular pronunciation of the Classical Latin cluster /kt/.
    In fact, we have act, efectiv (Romanian), acte, effectif (French), acte, efectiu (Catalan), acto, efectivo (Spanish, Portuguese before acordo ortográfico).

    In your example, detective and directory are borrowed from English, they are foreign words whose spelling is unalterated, factotum and cactus are non-translated Latin words (like forum), they retain the um and us endings, -ectomia is a specialistic suffix from the Greek.

    Another example is stretto: estricto/estrecho/estreito (Spanish, Portuguese before acordo ortográfico), strict (Romanian), strict (French), estricte/estret (Catalan).

    In Italian both meaning ("strict" and "thin") are stretto, while in the other language we have a learned form for the first meaning and a popular form for the second one.

    It is evident that there were two different strategies between Italian and the other Romance languages.
    Last edited: Sep 26, 2015
  28. Villeggiatura Senior Member

    Can you list the exceptions in Italian?
    I know one, perhaps two:
    practica - pratica
    cybiosactes - cibiosacte? cibiosatte?
  29. bearded

    bearded Senior Member

    Hi, Villeggiatura
    I think that there are several, but the following is just occurring to me:
    Arcticu(m) - artico.
  30. Nino83 Senior Member

    I have a master in law and I've never read practica, but only pratica.

    Cybiosactes, cibiosacte, cibiosatte doesn't exist in Italian.

    Hi bearded man. Where have you read this word?
    It is not present in vocabolario Treccani.

    Note: I'm speaking of Romance words, not of words retaining the -um and -us endings (which are Latin) or borrowed from English without any spelling change.

    EDIT: ah, you mean arcticum > artico, ok.
    Last edited: Sep 26, 2015
  31. Villeggiatura Senior Member

    I mean from practica (Latin) to pratica (Italian)
    ct only please, no nct rct, no ct+i-diphthong
  32. bearded

    bearded Senior Member

    Well, 'artico' and 'antartico' are very common Italian words, and the absence in Treccani surprises me.
    Artico from 'arcticu(m)' from Greek adjective arktikòs (originally ''under the constellation of the bear/arktos'') relating to the North Pole.
    Anyhow, apparently Villeggiatura does not accept examples of ''consonant+ct'', although to me 'artico'' seemed to be a good example of reduction of ct to simple t. But yes, he is right: the cluster rct is something different.
    Last edited: Sep 26, 2015
  33. Nino83 Senior Member

    Yes, bearded man, I didn't get the meaning of the symbol "-" instead of ">".

    Yes, rct is different.
    We're speaking of the evolution of intervocalic ct (not followed by yod), which became it in Western Romance languages, mantained in Portuguese but evolved in French (/t/ not pronounced and /i/ coalesced with the preceeding vowel) and Spanish (ch).

    About practica > pratica, it derives from Greek and it is a proparoxytone word, so the stress could have influenced the normal evolution.
    In Latin words ct is often present in paroxytone words, if we exclude those words formed with suffixes, for example -arius, where by analogy with the simple word, the evolution was the same, for example refractus, refractarius > refrattario (also this is not a popular word, it is learned).
    Last edited: Sep 26, 2015
  34. bearded

    bearded Senior Member

    Then how do you explain 'eclettico' which is also proparoxytone and also derived from Greek (eklektikòs)? Why is it not 'eclètico'?
  35. Nino83 Senior Member

    The word practica derives from the verb practico, so -ctica is in the stem while eklektikos derives from ἐκλέγω, eklego, so it could be that in this case -cticus was seen as a suffix, like in refrattario.
    This is only a possible explanation, I'm not sure of it and I don't strongly propose it. :)

    It can be that pratica is an exception. But, also in this case, the Classical Latin pronunciation /kt/ was not retained, while the other Romance languages retained it.
    Last edited: Sep 26, 2015
  36. bearded

    bearded Senior Member

    That is only true if 'pratica' is a verb (3rd person), but not if it is a noun or an adjective, as it can well be in our language. Like eklekticos, also praktikos is from a verb (prasso < prak-so) there should be no real difference.
    I think that 'pratico/a' is a real exception, as I do not see why the pronunciation ct>tt has not been retained.
    ''Val meglio la pratica della grammatica'':)
  37. Nino83 Senior Member

  38. Penyafort

    Penyafort Senior Member

    Catalan (Catalonia), Spanish (Spain)
    Catalan reduced -IT- into -T-.

    FACTU [faktu] > [fajtu] > feyt > fet /fet/
    LACTE [lakte] > [lajte] > lleit > llet /ʎet/
    LÈCTU [lɛktu] > [lejtu] > llit /ʎit/
    PÈCTUS [pɛktus] > [pejtus] > pits > pit
    STRÌCTU > [strejtu] > estret
    DICTU > [diktu] > [dijtu] > dit
    NÒCTE [nɔkte] > [nojte] > nuit > nit
    [lukta] > [lujta] > lluita /'ʎujtə/

    But it was preserved if at the beginning:
    ÒCTO [ɔkto] > [ojto] > uit > vuit, (Valencian) huit
  39. bearded

    bearded Senior Member

    Re: (Italian) pratico < practicu(m)
    My theory is that the c was dropped to form a word with -atico ending in analogy with numerous other adjectives - of Latin and Greek origin - having the suffix -atico.
    Simpatico estatico selvatico pratico.
  40. Nino83 Senior Member

    In popular words, but in learned words it follows the French, Spanish, Portuguese, Romanian strategy, i.e it retains the Classical Latin pronunciation and spelling.

    Some passage (diphthong) missing here.

    Convincing, it could be.
    Last edited: Sep 26, 2015
  41. robbie_SWE

    robbie_SWE Senior Member

    Trilingual: Swedish, Romanian & English
    I would like to add an exception to Danielstan's post about the Romanian evolution of Latin -ct-.

    a vătăma ("to wound, to injure" < Latin victimāre) -> something similar happened in Portuguese vitimar.

    Although not a part of this discussion, it is interesting to extend the discussion to why Latin -pt- also underwent significant changes in Romance languages.

    Latin septem / ruptus / neptis / tempto:

    Catalan - set / - / néta, nét / tentar
    French - sept* / - / nièce / tenter
    Italian - sette / rotto / - / tentare
    Portuguese - sete / roto / neta, neto / tentar
    Romanian - șapte / rupt / - / -
    Spanish - siete / roto /nieta, nieto / tentar

    * (N.B. silent "p")
  42. Testing1234567 Senior Member

    Hong Kong
    Seeing that French seldom mentioned here, I'm trying to throw in my little bit of knowledge in here:
    FACTU [faktu] > [fajt] > [fɛt] > fait /fɛ/
    LACTE [lakte] > [lajt] > [lɛt] > lait /lɛ/
    LÈCTU [lɛktu] > [ljɛjt] > [lit] > lit /li/
    PÈCTUS [pɛktus] > [pjɛjts] > [pis] > pis /pi/
    TÉCTU [tektu] > [tejt] > [toit] > toit /twa/
    DIRÉCTU [direktu] > [dəreit] > [droit] > droit /dʁwa/
    STRÌCTU [striktu] > [(ɛ)strejt] > [ɛstroit] > étroit /etʁwa/
    DICTU [diktu] > [dejt] > [dit] > dit /di/
    NÒCTE [nɔkte] > [nwɔjt] > [nyjt] > nuit /nɥi/
    FRUCTU [fruktu] > [fryjt] > [fryjt] > fruit /fʁɥi/

    OCTO [ɔkto] > [wɔjt] > [yjt] > huit /ɥit/


    EDIT: inserted "DIRECTU > droit"
    Last edited: Sep 26, 2015
  43. Testing1234567 Senior Member

    Hong Kong
    SEPTE [sɛpte] > [sɛtt] > sept /sɛt/ ("p" is a learned insertion)
    NEPTIS > (V.L.) NEPTIA [nɛptja] > [njɛttʲə] > [njɛsə] > nièce /njɛs/
    TEMPTARE [temptare] > [tẽmtær] > [tẽtɛr] > tenter /tɑ̃te/
  44. francisgranada Senior Member

    If we can extend this discussion also to -pt, then in modern Italian we have opzione and optare, that "normally" should have become *ozione and *ottare.
  45. Testing1234567 Senior Member

    Hong Kong
    More "pt" evolution to French:

    SCRIPTU [skriptu] > [(ɛ)skritt] > [ɛskrit] > [ɛkrit] > [ekrit] > [ekri] > écrit /ekʁi/
    SUPTUS [suptus] > [sots] > [sous]> [su] > sous /su/

    EDIT: corrected some steps in "SUPTUS>sous"
    Last edited: Sep 27, 2015
  46. Testing1234567 Senior Member

    Hong Kong
    This can be easily explained by "they are borrowed after the pt>tt change".
  47. danielstan Senior Member

    Romanian - Romania
    Thank you!
    I have not known this exception and it escaped to some Romanian linguists, too (because the ct > pt phonetic rule is treated in many Romanian books).

    Another one is:
    strâmt ("narrow" < lat. *strinctus < Classical Latin strictus)
    I don't know how they reconstructed the *strinctus, but I guess this *strinctus could reasonably explain words in other Romance languages or dialects.
    I also know that strâmt was also spelled strimt in some Romanian texts of 16th century - probably it was pronounced like this in some regions with dialectal differences.

    Out of topic: The online Romanian dictionary I use is: (DEX = dictionar explicativ (al limbii române))
    which uses printed dictionaries edited by the Romanian Academy.
  48. Testing1234567 Senior Member

    Hong Kong
    French étreint < estreint < streint, which only correspond to strínctus, because *strentus would give *étrent
    Italian strinto < strintus, which can also correspond to strínctus
  49. Testing1234567 Senior Member

    Hong Kong
    Or by comparing with other words.

    Latin cinctus > CINCTU [kinktu] > [kʲeint] > [seĩt] > ceint /sɛ̃/
    Latin tinctus > TINCTU [tinctu] > [teint] > [teĩt] > teint /tɛ̃/
    Latin punctus > PUNCTU [punktu] > [point] > [poĩt] > point /pwɛ̃/

    (Oops, all of them are French.)
  50. Testing1234567 Senior Member

    Hong Kong
    (I know this is off-topic)

    It's fun to predict what CORRECTUS would become if it went through the dark ages.

    CORRÈCTU [kɔrˈrɛktu] > [kərˈrjɛjt] > [krjɛjt] > [krit] > *crit /kʁi/

Share This Page