Wikidata:Lexicographical data/Sandhi rules

The intent of this page is to list some examples of different sandhi rules, morphophonemic tendencies, which govern the pronunciation of various lexeme forms, or determine which lexeme form is used. This may help to determine how to model these.

English

• a (L2767)

Rule: For F1 and F2…

• If followed by a word with an initial consonant phoneme, F1 must be used, written as a and spoken as ə (with some variation for stress)
• If followed by a word with an initial vowel phoneme, F2 must be used, written as an and spoken as ən (with some variation for stress)

Indication in writing is mandatory and always follows pronounciation tendencies regardless of differences in orthographic representations.

• Plural suffix -s
  • [ɪz] after sibilant consonants
  • [s] after voiceless consonants, [z] after voiced consonants and vowels

• in- (L35577)
  • ir- before [r], il- before [l], i- before [g], im- before nasal consonants

• the (L2768) (e.g.)
  • [ðiː] when stressed
  • [ði] when unstressed before vowels
  • [ðə] when unstressed before consonants

Maltese

• il- (L314589)
  • il- before non-coronal consonants, the "l" assimilates before coronal consonants. This also applies to preposition+article combinations.
  • il- (etc) at the beginning of a sentence and after consonants, l- (etc) after vowels.

Mandarin

• 一 (yī)
  • 4th tone (yì) when followed by 1st, 2nd or 3rd tone

• 不 (bù)
  • 2nd tone (bú) when followed by 4th tone

Mongolian

Mongolian has vowel harmony.

instrumental case suffix

• -аар (back unrounded)
• -ээр (front unrounded)
• -оор (back rounded)
• -өөр (front rounded)

ablative case suffix

• -аас (back unrounded)
• -ээс (front unrounded)
• -оос (back rounded)
• -өөс (front rounded)

question particle

• уу (back)
• үү (front)
• юу (back, after vowel)
• юү (front, after vowel)

Prakrit

Verb stems undergo certain sound changes when attached to certain affixes. See Prakrit Morphology for more details.

• In the present, first person singular indicative forms, stems ending in -a are lengthened to ā. In the first person plural inflection in this paradigm, the same vowel may either change to -ā or -i.
• In the second person singular imperative forms, stem-final -a is changed to -ā if followed by the affix -hi. Note that there are two alternate affixes for this form (zero affix or -su) which do not lengthen the stem vowel.

Punjabi

• ਘਾਹ/گھاہ (L685963)

Rule: For F1…

• If followed by a word beginning with a level tone syllable, form is monosyllabic low tone /kɑ̀/
• If followed by a word beginning beginning with a high tone or low tone syllable, or if spoken in a single-word sentence, form has a second syllable consisting of the inherent vowel with level tone /kɑ̀.ɐ/

This pronunciation distinction is never indicated in spelling.

• ਚੜ੍ਹ/چڑھ (L716995)

Rule: For F1…

• If followed by a word beginning with a level tone syllable, form is monosyllabic high tone /t͡ʃɐ́ɽ/
• If followed by a word beginning with a high tone or low tone syllable, or if spoken in a single-word sentence, final consonant of form is spoken as separate syllable with vocalic release (micro-schwa) /t͡ʃɐ́.ɽᵊ/

This pronunciation distinction is never indicated in spelling.

• ਵਿਚ/وِچ (L679728)

Rule: For F2, F3/F7 (spelling variants), and F9…

• If preceded by a word for which the final syllable boundary is not a consonant cluster, _and_ the word-final phoneme is not tʃ ~ چ ~ ਚ or /d͡ʒ/ ~ ج ~ ਜ, contraction F3/F7 is typically used, indicated in writing as چ ~ ਚ, and pronounced as /t͡ʃɐ/
• If preceded by a word for which the final syllable boundary is not a consonant cluster, _and_ the word-final phoneme is not a vowel, contraction F9 may be used, written as اچ ~ ਇਚ and pronounced as /ɪt͡ʃ/. Effectively, this is typical for words ending in the consonants excluded by the above rule. Anecdotally, while this distinction is apparent in speech, it is much more commonly distinguished in Shahmukhi writing than in Gurmukhi. (\* This is likely attributable to this form requiring more pen strokes than the full form in the latter script.)
• If preceded by a word for which the final syllable boundary is a consonant cluster, the full uncontracted form F2 must be used, spelled وچ ~ ਵਿਚ and pronounced /wɪt͡ʃ/

Technically the uncontracted form may be used in any context, and in more conservative or formal writing this is done. However, "typical" speech and writing does employ contractions for this lexeme for metrical/rhythmic reasons.

---

• verbal participle suffix ݨا ~ ਣਾ

• after retroflex consonant, نا ~ ਨਾ
• after any other phoneme, ݨا ~ ਣਾ

Welsh

• y/yr/'r
  • 'r after a vowel
  • yr before a vowel, y otherwise