Generally, Mood is marked by suffixes, prefixes, infixes etc. But are there languages which have mood marked by phonetic changes at syllable / word boundaries?
In English, mood is marked by the modal auxiliary verbs.– BillJApr 3, 2018 at 8:10
yeah but are there languages with mood marked by phonetic changes?– WiccanKarnakApr 3, 2018 at 8:21
2Apparently one class of verbs in Saho has vowel changes for certain moods, e.g. eerhege "I know" vs. aarhago "that I know" (subjunctive mood), or (in Irob) igriʕ- "I cut" (indicative) vs. agraʕ- "I shall cut" (jussive). Is that the kind of thing you're looking for? I understand "phonetic changes" to mean morphological apophony, but I don't get the "syllable/word boundaries" part.– melissa_boikoApr 3, 2018 at 10:01
2That sounds a bit difficult to distinguish as sound change/apophony vs. affixes. Suppose you have some language where, say, pataka is "I run [indicative]" and bataka is "I run [subjunctive]". You could say that the subjunctive voices the /p/, but you could equally say that the root is -ataka and p-/b- are IND/SUBJ modal prefixes. The only way I can think of them being undeniably not prefixes is if some feature (like voicing or vowel backing etc.) was applied consistently to different phonemes (like the Jap. rendaku you mentioned); but I've never heard of a language with featural mood.– melissa_boikoApr 3, 2018 at 10:43
1@WiccanKarnak Yes, similar to that, but English only does it for tense, whereas Semitic languages do it for modality too.– curiousdannii ♦Apr 6, 2018 at 3:52
Yes or no, depending on what exactly you are looking for. In most Bantu languages, tense etc. inflection for verbs involves adding certain tones. Verb inflection in Bantu covers all sorts of things, including mood, and there are indeed tonal markings that indicate subjunctive and imperative. This is frequently realized as a particular tone pattern on the last vowel of the verb (at the rightmost word boundary). I'm not sure that that constitutes a "change", since all of the tonal distinctiveness is on the root-initial syllable, and the tones added for inflection really constitute more of a choice than a change. Also, it's not really phonetic, it is phonological (these are not low-level adjustments in H and L tones, these are categorial distinctions).
In Shona, there is an actual change which you could say marks a mood: interrogatives can be marked by changing the last H tone of a word to L. This is indeed a change, from whatever tone the final vowel would have (by whatever set of rules) to L, and this is the last rule of the grammar. However, this applies to any word, and I would say is not a "marker" in the usual sense, it is a phonological rule that applies in a specific pragmatic context that happens to correspond to a mood as found in some inflectional systems. Also, this applies at utterance boundaries, not word boundaries.
In Gurage languages, there are palatalizations and labializations that indicate various categories, and which affect root consonants. The labialization effect can apply to multiple consonants (non-coronal sonsonants can be labialized). Because Gurage has no C-clusters longer than 2, all consonants are at a syllable boundary. However, I don't believe that the labialization prosody happens to mark mood.
German marks mood by umlaut, e. g.
ich mag (indicative) > ich möge (subjunctive II)
ich sehe (>preterite: sah) > ich sähe
ich liege (> ich lag) > ich läge
ich trage (> ich trug) > ich trüge, etc.
(a + j > ä /ɛ/, u + j > ü /y/,...)
This inflection is based on i-umlaut motivates different morphophonological developments. In older German the non-real moods were marked by suffigation of a /-j/, but this suffix was incorporated into the verb stem. Back vowels become fronted by merging with the /-j/, front vowels are moved towards /e/. This development is similar to ablaut which is basically used to mark tenses, namely present > preterite > perfect participle (i > a > u) with falling vowel formant gradient. Since tense is never purely tense but also mood (i. e. accessibility of the event from the context at hand) preterite served as a base for the less immediate, less accessible irrealis expressed by subjunctive II.
(NB causatives were also derived by i-mutation, e.g. food > feed in Engl., trinken (“to drink”) > tränken (“to make/let someone drink”)
I wouldn't say umlaut is a phonetic change at the syllable/word boundary ... in fact, the characteristic feature of umlaut seems to be that it is a word-internal vowel change. But I guess looking back at it, I don't really understand what the question means by "phonetic changes at syllable / word boundaries". Apr 5, 2018 at 17:04
I did add a rendaku example in the comment @sumelic Apr 6, 2018 at 8:31