Cognitive morphology in finnish: Foundations of a new model

Abstract

We summarise the main results from a series of Finnish studies dealing with single-word experiments with aphasics as well as lexical decision and eyemovement registration tests performed on normals. On the basis of our experimental results, we propose a processing model of Finnish nouns. For the input and central lexicons, this Stem Allomorph/Inflectional Decomposition (SAID) model assumes morphological decomposition of inflected (with the exception of the most frequently encountered inflected noun forms) but not derived noun forms. For the output lexicon, it predicts that both inflected and productive derived forms have decomposed representations. In the case of marked stem variation (resulting from stem formation and/or morphophonological alternation), the model assumes that the stems are represented by their allomorphs, and not by a single morph. In this respect, our model postulates more suppletion in the input/output lexicons than would be predicted on the basis of formal morphological analyses. However, among the allomorphs, the nominative singular of nouns appears to have a special status.

Affixal salience and the processing of derivational morphology: The role of suffix allomorphy

“…Whereas the evidence suggests that Finnish inflected words tend to be morphologically decomposed (Laine, Vainio, & Hyö nä , 1999;Niemi, Laine, & Tuominen, 1994), the recent empirical evidence from Finnish derived words speaks for the opposite. Niemi et al (1994) proposed that Finnish derived words are recognised holistically in normal comprehension.…”

“…Whereas the evidence suggests that Finnish inflected words tend to be morphologically decomposed (Laine, Vainio, & Hyö nä , 1999;Niemi, Laine, & Tuominen, 1994), the recent empirical evidence from Finnish derived words speaks for the opposite. Niemi et al (1994) proposed that Finnish derived words are recognised holistically in normal comprehension. For supporting evidence, Laine, Niemi, Koivuselkä -Sallinen, and Hyö nä (1995) found that the performance of an aphasic deep dyslexic patient was significantly better for derived than for inflected words both in single word-reading tasks and in a repetition experiment.…”

Three factors in language variation

“…Though direct storage of derived forms has been suggested (Baayen 2003), 16 the combinatorial explosion of morphologically complex languages (Hankamer 1989, Niemi, Laine, & Tuominen 1994cf. Chan 2008) necessitates a stage-based architecture of processing that produces morphologically complex forms by rule-like processes (Caramazza 1997, Levelt et al 1999.…”

Prosody-assisted head-driven access to spoken German compounds.

“…These models postulate two possible routes of identifying complex words: (a) a decompositional route and (b) a direct route in which the word as a whole is accessed by a lexical entry. Either these two routes operate in parallel (Caramazza, Laudanna, & Romani, 1988;Frauenfelder & Schreuder, 1992;Schreuder & Baayen, 1995), or their activation depends on the type of complex word (see the nonparallel model proposed by Niemi, Laine, & Tuominen, 1994).…”

“…Conversely, compound words with an opaque head are identified by a whole-word process. Because neither nonparallel models (the stem allomorph/inflectional decomposition [SAID] model, Niemi et al, 1994) nor parallel models (the augmented addressed morphology [AAM] model, Caramazza et al, 1988; the morphological race model, Frauenfelder & Schreuder, 1992; the meta-model, Schreuder & Baayen, 1995) seem to be able to account for the strong distinction emerging from our data, we formulated the following explanation.…”