Kari Suomi

Durational and tonal correlates of accent in Finnish

Abstract The paper reports the first study that explicitly distinguishes the phonetic correlates of sentence accents from those of word stress in Finnish (stress acted as the baseline whose correlates as against of stress were not investigated). Sentences were constructed that successfully elicited no accent, moderate accent, and strong accent on target words. The three degrees of prominence were clearly differentiated phonetically. Mere word stress was not signalled tonally, while accents were signalled mainly tonally. Strong accent involved longer segmental durations than the other degrees of prominence. Timing and extent of the accentual tonal movements were highly uniform across different word structures, and not invariably tied to the initial (stressed) syllables. The data were consistent with a view that the timing of f0 movements was dependent on the moraic structure of target words. This finding seems to be connected with the typologically rare combination of stress (or rhythmic) and quantity systems in the language.

Utterance-final lengthening and quantity in Northern Finnish

Abstract Utterance-final lengthening in Northern Finnish was investigated using tightly controlled laboratory materials, with particular focus on its interaction with the languages single (short) vs. double (long) vowel distinction. Like many other languages, Finnish exhibited utterance-final lengthening, although the estimates of magnitudes of lengthening on final vowels varied greatly depending on the treatment of the utterance-final breathy/voiceless portion of the vowel. As has been also shown for other languages, the lengthening occurred as early as the stressed, penultimate syllable of disyllabic words and was generally progressive. Crucially, vowel quantity interacted with the lengthening in a manner consistent with a hypothesis that Finnish regulates utterance-final lengthening to preserve its quantity system. Specifically, the voiced portion (the portion that is relevant to the perception of vowel quantity) of the longest single vowel (the half-long vowel) was restricted. Additionally, double vowels were lengthened less when the vowel in an adjacent syllable was also double, suggesting syntagmatic constraints. Our results support the view that utterance-final lengthening is a universal tendency but is implemented in language-specific ways and must be learned.

On the tonal and temporal domains of accent in Finnish

Abstract Using monosyllabic words that can be continued to quadrisyllabic words (for example, sei, Seiko, Seikola, Seikolasta), all spoken with two degrees of prominence (unaccented and strongly accented), this study examined the temporal and tonal domains of accent in Finnish. Large accentual lengthening was observed to extend from word onset to the end of the third syllable, with minor lengthening appearing on the first segment of the fourth syllable. The tonal domain of accentuation in turn was observed to extend from word onset to the middle of the third syllable, and in shorter words, to a corresponding temporal location in the next word. Thirdly, it was observed that polysyllabic shortening does not operate in Finnish: word length (number of constituent syllables) has no overall effect on segment durations. The results, together with previous ones, show that in Finnish, a full-fledged quantity language, segment durations are adjusted to achieve a temporally and tonally uniform realization of accent. This is contrary to the situation in many nonquantity languages, in which the temporal realization of accent varies as a function of the segmental structure of the accented syllable.

On durational correlates of word stress in Finnish

Abstract It has been shown that the phonetic alignment of (sentence) accentuation in Finnish follows a moraic pattern: an f0 rise during a words first mora, a fall during the second one. This paper shows that the alignment of Finnish (word) stress, which is fixed and associated with the initial syllable, is similarly realized over two morae. It is shown that segments and syllables have a reliably longer duration when they occur within the domain of the words first two morae than when they occur outside this domain. The most conspicuous concomitant of primary stress is the durational variation of the second-syllable single vowel: if this vowel constitutes the words second mora, its duration is very much longer than otherwise. Secondary stress is shown to involve a similar although attenuated durational variation in the second-syllable vowel of the words second foot. It is further shown that when a consonant constitutes the words second mora, the consonant is lengthened, but only if it is voiced. It is argued that the motivation for the second-mora lengthening (applying as it does to vowels and voiced consonants) is to provide room for the phonetic realization of accentuation.

Temporal conspiracies for a tonal end: Segmental durations and accentual f0 movement in a quantity language

Abstract This study investigated the interaction between segment durations and the tonal manifestation of accent in Finnish, a language with a binary quantity opposition in both vowels and consonants, and fixed initial stress. Unaccented and strongly accented versions of four sets of disyllabic target words were elicited. Accentual lengthening affected all segments in the target words reliably, but the lengthening was highly nonlinear. Word-initial consonants were extensively lengthened, and otherwise the extent of lengthening was determined by whether or not a segment constituted one of the words first two morae, irrespective of syllable structure. The words first two morae were extensively lengthened, other segments less. Usually, those segments were lengthened in the accented versions that had relatively long duration in the unaccented versions. In some positions, the extent of lengthening seemed to be constrained by a need to maintain the quantity opposition. The accentual f0 rise-fall movement was highly uniform across the four target word sets, and the author argues that the observed subphonemic durational variations, besides making room for the tonal movement, conspire to guarantee tonal uniformity across the different word structures, all consisting of disyllabic feet.

Palatal Vowel Harmony: A Perceptually Motivated Phenomenon?

The paper attempts to determine the motivating causes of palatal vowel harmony (PVH). Previously suggested causal explanations of PVH are critically evaluated, especially the progressive palatal assimilation view of the origin of PVH. Data on PVH restrictions from Finnish and Turkish are examined against a set of perceptually motivated working tendencies. It is shown that PVH and “labial harmony” are only special cases of a single, unitary type of restriction, statable acoustically in terms of the frequency of the second formant. A perceptual theory of the causes of PVH is proposed and general phonetic conditions for and against the development of PVH in a language are suggested. On the whole, the paper is an argument for a substance-based approach to phonology.

Quantity constraints on the temporal implementation of phrasal prosody in Northern Finnish

This study investigated interactions between vowel quantity and two types of prosodic lengthening (accentual lengthening and the combined effect of accentual and utterance-final lengthening) in disyllabic words in Northern Finnish. Two quantity-related constraints were observed. First, in both types of prosodic lengthening, vowels were lengthened less when they were next to a syllable containing a double vowel than when they were next to a syllable containing a single vowel (a quantity neighbour constraint). Second, a durational ceiling effect was observed for the phonologically single, half-long vowel under the combined effect of accentual and utterance-final lengthening. These findings can be seen to support the view that quantity languages regulate the non-phonemic use of duration because of the high functional load of duration at the phonemic level. Additionally, the combined effect of accentual and utterance-final lengthening appeared to have its own lengthening profile, distinct from the simple sum of the two lengthening effects suggested previously. Implications for speech timing research will be discussed.

Durational elasticity for accentual purposes in Northern Finnish

Abstract This study examined the elastic behaviour of segment durations in Northern Finnish CVC 2 V and CVC 2 C 3 V nonsense words produced in a constant frame sentence. The identities of C 2 and C 3 were systematically varied, and the identities of the other segments were fully counterbalanced. In the CVC 2 V words, the three consonants in the C 2 position exhibited systematic differences in their intrinsic durations, whereas in the CVC 2 C 3 V items they did not, and their durations were reliably longer than in the CVC 2 V items. The two consonants in the C 3 position in the CVC 2 C 3 V items also exhibited an intrinsic durational difference. For the (phonologically single) vowels in the target word structures, four complementary degrees of vowel duration were observed, with a distribution determined by the vowels moraically defined structural position in the word. Where consonants exhibited intrinsic or other durational differences, these differences were usually counteracted by compensatory shortening of another segment in the word. A locus of duration-to-tone adjustments is postulated that consists of the words first mora and the next two segments one of which is the words second mora. As a result of the elastic behaviour, the locus tends to have constant duration across different word structures, and it ensures the tonally and temporally uniform realisation of accent, a characteristic that seems to distinguish Northern Finnish from many other languages. These and other, previous durational findings in (Northern) Finnish are related to a recent model of speech timing.

Durational patterns in Northern Estonian and Northern Finnish

Abstract Estonian and Finnish are closely related languages in which quantity is extensively exploited for lexical and grammatical purposes (in both consonants and vowels, independent of each other), yet with several phonological differences between the quantity systems (e.g. a ternary opposition in Estonian, a binary one in Finnish). To date, segment durations in the two languages have not been systematically compared. This paper reports a necessarily explorative experiment with two primary goals: first, to compare the phonetic realisation of quantity in the two languages in selected word structures, and second, to relate the results on accentual lengthening to the predictions of a speech timing framework ( White, 2002 ) that has been developed on the basis of observations in an essentially non-quantity language, namely English. It was observed, besides cross-language differences in the durational realisation of the three-way and two-way quantity contrasts that durationally C1, outside the quantity system in both languages, behaves differently in the two languages. It was also observed that the patterns of accentual lengthening in the two languages are highly consistent with predictions of the timing framework.

The voicing contrast in Fenno-Swedish stops

Abstract This paper presents the results of an investigation of the fortis–lenis contrast in Fenno-Swedish stops. The data show that in utterance-initial position, the two-way contrast is realised as a contrast between unaspirated and prevoiced stops. Word-medially and finally, the contrast is that of voiceless, unaspirated stops and fully voiced stops. Fenno-Swedish is thus another Germanic language, like Dutch, Afrikaans and Yiddish, that has a contrast between plain unaspirated and prevoiced stops in utterance initial position. The stop contrast in Fenno-Swedish differs from that of Central Standard Swedish, in two ways: (i) In Central Standard Swedish the contrast is between aspirated and prevoiced stops in utterance initial position, whereas it is between voiceless, unaspirated stops and prevoiced stops in Fenno-Swedish; in medial and final position, one series of stops in Central Standard Swedish is voiceless (aspirated, preaspirated or unaspirated) and the other series is fully voiced, whereas in Fenno-Swedish, one series is voiceless and unaspirated and the other voiced, i.e., in Fenno-Swedish there is no aspiration. (ii) Variation was observed for some Fenno-Swedish speakers in the production of /b d ɡ/, with many tokens being completely voiceless and overlapping phonetically with tokens of /p t k/, whereas there is no overlap between the two stop categories in Central Standard Swedish. Stops in Finnish are voiceless and unaspirated. The fact that the aspirated stops have been lost in Fenno-Swedish, and the fact that there are voiceless occurrences of /b d ɡ/ tokens in Fenno-Swedish suggests influence from Finnish. Fenno-Swedish thus appears to pattern with other languages influenced by a language with a different system of laryngeal contrasts.