From Wiktionary, the free dictionary
Jump to navigation Jump to search
en-4 This user speaks English at a near-native level.
fr-3 Cet utilisateur peut contribuer avec un niveau avancé de français.
la-3 Hic usuarius probe ac latine conferre potest.
ka-3 ამ მომხმარებელს შეუძლია წვლილი შეიტანოს ქართული ენის ცოდნის მაღალ დონეზე.
uk-2 Користувач може робити внесок українською мовою на середньому рівні.
Also I can read—but not properly speak—the other major Italo-Western Romance languages.

I mainly deal with entries in Romance languages and (when relevant) Latin.


  • Recategorize all 1099 Catalan words that are claimed to be ‘inherited from Old Occitan’ as either inherited from Old Catalan/Latin or borrowed from Old Occitan.
  • Remove hundreds of fake Mozarabic words added by Romandalusí, often through thirty or so different IP's.
  • Sort through categories for Romance words claimed to be ‘inherited from Latin’ and remove the ones that clearly weren't.
  • Move all 'Reconstructed Latin terms' which are in fact attested (in the appropriate time period) to the mainspace, with citations.
  • Recategorize all ‘Terms inherited from Medieval Latin’ as one of the following: (largely complete)
    • Inherited from Early Medieval Latin (attested up to ca. 10th c. AD)
    • Borrowed from (later) Medieval Latin (attested 11th c. and later)
    • Inherited from a reconstructed ‘Vulgar Latin’ term (unattested anywhere prior to 11th c.)
  • Neapolitan clean-up:
    • Fix/check metaphonic plurals
    • Fix IPA transcriptions (largely complete)
    • Relemmatize verbs to ⟨-re⟩ spellings
  • Franco-Provençal overhaul:
    • Move all lemmas to ORB spellings
    • Add Swadesh list words
    • Add pronouns (largely complete)
    • Fix the verb conjugations
    • Add references to AIS/ALF/FEW wherever applicable
    • Add altforms per DFP/LTA/FEW



Phonological pet peeves

  • If a language has phonemic stress, its presence or absence should be indicated in monosyllabic words, not omitted for the sake of typographical convenience. Consider the minimal pair Anne :: an, that is /ˈæn/ :: /æn/. Representing both as /æn/ would fail to account for the differences in surface realization.[1]
  • There is no such thing as contrastive syllable division or contrastive secondary stress in any of the languages with which I am familiar.[2] Any claimed example falls apart once one accounts for morphology.
    • Night-rate :: nitrate = ⫽ˈnaɪt+ˈɹeɪt⫽ :: ⫽ˈnaɪtɹeɪt⫽[3] > [ˈnaɪt.ˌɹeɪt] :: [ˈnaɪ.ʧɹeɪt]. In other words, the reason they sound different is that the former (but not the latter) has /t/ and /ɹ/ split across two different morphemes, night-rate being a transparent combination of night and rate.
    • Reagan :: raygun = ⫽ˈɹeɪɡən⫽ :: ⫽ˈɹeɪ+ˈɡʌn⫽ > [ˈɹeɪ.ɡən] :: [ˈɹeɪ.ˌɡʌn]. That is, the latter (but not the former) is a transparent combination of two nouns, ray and gun, each with their own underlying stress.


  1. ^ Anne-Marie [ˌanməˈɹiː], an artist [ə̆~ăˈnɑːtɪst].
  2. ^ Nor, I suspect, in any human language.
  3. ^ One could also analyse nitrate as a confix of ⫽ˈnaɪtɹəʊ-+-eɪt⫽, with nitro- having the prevocalic allomorph nitr-.