Category talk:Mozarabic language

RFV discussion: May 2022

The following information has failed Wiktionary's verification process (permalink).

Failure to be verified means that insufficient eligible citations of this usage have been found, and the entry therefore does not meet Wiktionary inclusion criteria at the present time. We have archived here the disputed information, the verification discussion, and any documentation gathered so far, pending further evidence.
Do not re-add this information to the article without also submitting proof that it meets Wiktionary's criteria for inclusion.

---

Various Mozarabic terms

Follow-up from a previous section (related discussion here and here). For a couple days I've been cleaning up the Mozarabic entries here in Wiktionary: moving them to the attested spelling(s), adding quotations, and changing the transcriptions to the ones used by the sources. However, there are a few words which I haven't found in either {{R:mxi:CorrienteHeb}} (for texts in the Hebrew script) or {{R:mxi:Jones}} (in the Arabic script). As far as I know, these sources cover the whole Mozarabic corpus (which is rather small); I could have missed some words, but others seem completely made up. This is a list of the words I haven't found in the corpus:

• bont, إسقللايرة (escalâira), العجمية (alǧamiah), الغودون (alǧodon), اِيّارِ- (-iyyáre), شّوشّْ (šoš), قاذة (káda), قرآن (qurán), ليطه (laitah), مَطْرُينة (matrwéna), يَنّايْرْ (yanayir)
• These terms have a quotation, but I suspect it's made up (at least the spelling used for it certainly is): شوقّه (súkko), ليُ (liyu)
• These aren't attested but can be reconstructed (they appear in {{R:DAAL}} with an asterisk): البيست (alpîšt), جّيمْجِّ (čímče), شّولُ (šúlo)
• Made-up spellings of attested terms, fanciful transcriptions into Arabic of romanizations: طو (tu) (attested as ت (tu)), عَان (en) (attested as אן (en) and ־ן (-n)), ايلجو (welyo) (attested as ולויש (wélyoš))

Santi2222 (talk) 23:02, 6 May 2022 (UTC)

Hi @Santi2222

• ليطه (laitah), a purported derivative of Latin laetus, is without a doubt made-up and can be promptly deleted. The user Romandalusi read that Mozarabic conserves the diphthong /ai̯/ and wrongly assumed that meant a continuation of the Latin /ae̯~ai̯/ rather than a preservation of early Romance /ai̯/, which never derives from the Latin diphthong (cf. Italian amai 'I loved', fornaio 'baker' < Latin amavi, furnarius, not *amae, *furnaeus).
• Searching 'escalaira' on Google Books brings up a fair number of results, including apparently Corriente 1992. I don't have access to the entire book, so I can't determine what the ultimate source of this word is. It could be from something other than a kharja.
• The quote justifying شوقّه (súkko)} and ليُ (liyu) appears to be a questionable reading of kharja 25:

amanE, ya habib,/ alwahsha ME+N FARAS!/ BEN, BEJA MA BOKELLA,/ LEW SUKKO TE+N BEBRAS' (source: could only find this on a random blog)

Jones (1988: 189) gives a completely different reading of the last part, which has neither of the words in question. Accordingly, I think it'd be best to remove both words from Wiktionary- the evidence is simply too flimsy.

• We should probably create a new category for reconstructed Mozarabic words, to make it clear that they aren't actually attested. The question is what sort of spelling to use for them. Perhaps Latin script would be best, if that is what the DAAL uses.
• As for the made-up spellings of attested terms, I see no reason not to promptly delete them. Nicodene (talk) 09:19, 9 May 2022 (UTC)

For the reconstructed terms, I think we should keep the DAAL spelling (making up a spelling in a script in which the language didn't have anything close to a standardized spelling seems out of the question to me). For example, moving شّولُ (šúlo) to something like *śúlo (in this particular case, it was borrowed into Andalusian Arabic xúlu, and then into Spanish). Regarding the words from Kharja 25, I agree that they should be deleted, the proposed reading seems too speculative (even for the standards of Mozarabic). I've also noticed that some of the supposedly Mozarabic terms are actually Andalusian Arabic (which has a way larger corpus). This is the case of يَنّايْرْ (yanayir), attested as AA يَنَّيْر (yannayr) (I'm trying to find a quotation for this one now).Santi2222 (talk) 12:05, 9 May 2022 (UTC)

No, “they appear in {{R:DAAL}}” does not compel us to conclude “they can be reconstructed”. Can’t just reconstruct, say, from the appearance of alpiste in Galician the presence in Portuguese; there also were direct contacts of Arabs with speakers of other Iberian Romance languages than Mozarabic. Keep up the focus on the attestations. And don’t cling to those bizarre transcriptions “used in the sources” (in fact by fanciful reference works; have I mentioned that on the Arabist side we are annoyed about the nonstandard transcriptions of Arabic words and innumerous links including the article الـ (al--)? The article is never part of the lemma, in the Andalusian Arabic dialect neither, and WT:AR TR applies very well, and no marking stresses is needed since they are not distinctive and the Romanists are wrong about them too, as I have shown on حَبَق (ḥabaq)). Fay Freak (talk) 14:31, 9 May 2022 (UTC)

But I see you already normalize spellings which have followed Pedro de Alcalá. If you go the full hog you normalize to Arabic spelling. I don’t think Andalusian Arabic should ever be present in Latin script; the other dialects are not either (though Modern South Arabian is), after all it is not the alphabet used by the language community (instead even Romance under Arabic rule was written in Arabic script; MSAL are unwritten). chírba to چِرْبَة, plural چِرَاب, also the consonants under which it is found in {{R:xaa:ELA|II}}. bedústar may point to بِيدُسْتَر (bīdustar), perhaps Andalusi بَيْدُسْتَر (baydustar), found in a less common spelling of جُنْدُبَادَسْتَر (jundubādastar) which is a usual term while this claimed term for “beaver” is not found but in Pedro de Alcalá, normally only this medical ingredient, so it can well be a ghost word introduced by Pedro de Alcalá by interpretation of جُنْدُبَادَسْتَر (jundubādastar) by the parts. He also uniquely claims قُسْطَاس (qusṭās) (Corriente cites him after his own edition page 245) to be the word for “scales”, though it isn’t otherwise used in the standards or dialects for “scales” but boils down to a Qurʾān interpretation, so demonstrably transmits his hearsay of the contents of the Arabic language rather than use that has existed in his era. كُرَّج (kurraj) also has been glossed as “colt” or “foal” (مُهْر (muhr)) by various dictionaries, interpreting it by their Persian knowledge, though in use it has the peculiar restriction of “hobby-horse” (this at least Corriente has not fallen for, for Vahagn wondering: it's the lectio difficilior behind rough glosses, as organisms use to be described not specifically enough in lack of detailled acquaintance, and other identifications from others are copied over for the same reason). Fay Freak (talk) 17:03, 9 May 2022 (UTC)

I feel like we are drifting a bit from the primary topic here, but anyway. I agree that words whose only attestation is from a sixteenth century glossary shouldn't be our focus here to start with. I don't like these entries and I haven't added any (I wouldn't mind if we excluded terms only found in the glossary altogether), unless you count moving this word. But if we're gonna have them, I don't think having them in an unattested spelling is a good idea. The "normalization" I mentioned in a changeset is basically substituting the háček-like diacritic in the original glossary by an acute accent. Given that the existing entries quoting the glossary already substituted the long ſ by the "normal" s, I don't see why we would want to keep that diacritic (either keep both the ſ-s and the ǎ-s or get rid of them, but not just one thing). In any case, I think this is more of a "normalized typography" rather than a "normalized spelling" thing, and not really an important matter (compared to keeping these entries or not). Otherwise, I totally agree that Andalusian Arabic entries should be in the Arabic script.Santi2222 (talk) 20:05, 9 May 2022 (UTC)

@Santi2222: To some degree, if the dictionary is supposed to be complete, reverse-transcription cannot be avoided. Modernly of course, cuneiform could not be printed well, and rarely is still. Gafat is probably attested more transcriptionally than in Ethiopic script, so some other now or soon defunct Ethiosemitic languages, and we are supposed to imagine how a local would write it rather than randomly creating entries as Latin and Ethiopic. This argument works for late medieval works as well, if we think the script is clear enough to show us the distinctive parts needed to spell natively. So Punic I have some terms where the spelling is comparatively clear but the only attestation is in Greek transcription; in some cases (one is farfaria) even Latin is only found in Greek transcription, in others a Greek term is mentioned as such by Latin authors (to be distinguished where a Latin term is an obvious Graecism but the Greek is not attested in neither use nor mention, e.g. molochitis and gossypium, as many terms in Pliny (you see with the latter previous philologists and a Wiktionary editor have been inexact)). The attestation situation of Polabian spellings and that of other extinct West Slavic languages is also readily circumstantial, yet somehow book-form dictionaries always end up having a uniform scheme, only the free-for-all nature of Wiktionary manages to convolute language documentation due to lacking prejudice for spelling systems.

This of course works more with Arabic than Mozarabic. In Mozarabic spelling there was no system, the Arabic script is fitted to Arabic and foreign to Mozarabic. And the Ethiopic Script is fitted to Ethiosemitic but in Cushitic we have all three of Geʿez script, Latin and Arabic one. So Romandalusí, insensitive to the distinctions of languages, made up Mozarabic spellings, in addition to ignoring the question of attestation altogether. (I mean yeah, like you listed it there were multiple layers of problems, rooted in a general lack of sensitivity, I roll it up this way.) Fay Freak (talk) 02:39, 10 May 2022 (UTC)