Processing modifier-head agreement in reading: Evidence for a delayed effect of agreement

Processing modifier-head agreement in long Finnish words: Evidence from eye movements

“…Our aim was to study whether word length regulates the time course of the processing of grammatical agreement by comparing the results of the current study with those of previous studies (Vainio et al 2003;2008), in which processing of grammatical agreement was studied for relatively short words.…”

“…Similarly, Vainio et al (2008) showed processing benefits for agreement when both modifier and head noun were in the inessive case (e.g., vanha/ssa talo/ssa 'old/in house/in' → 'in the old house'). Also here the effect showed up relatively late in the processing stream, such that word n + 1, the word following the target noun talo/ssa, was read faster when it was preceded by an agreeing modifier (vanha/ssa) than when no modifier was present.…”

“…We did so by constructing experimental materials and selecting experimental target words very similar to the materials of Vainio et al (2008), except that the target nouns in our study are of considerable length (14-16 characters), whereas the target words in Vainio et al were 7-8 characters long. In other words, we pitted long words preceded by an agreeing modifier against long words preceded by no modifier and investigated (a) whether there was a positive effect of agreement and (b) what the time course of such an effect would be in comparison to that observed in the Vainio et al (2008) study. To our knowledge, there are no previous studies that have investigated whether word length affects how modifier-head agreement is processed during reading.…”

“…The RT methods include a maze task (a modified word-by-word paradigm, Nicol et al 1997), naming (Guillelmon-Grosjean 2001) and lexical decision (Gurjanov et al 1985;Lukatela et al 1987;Schriefers et al 1998), whereas on-line methods include the registration of event related brain potentials (ERP) (Coulson et al 1998;Friederici et al 1999;Gunter et al 2000;Osterhout-Mobley 1995) and eye tracking studies (Deutsch-Bentin 2001;Pearlmutter et al 1999;Vainio et al 2003;2008). As both Gunter et al (2000) and Pearlmutter et al (1999) stress, by using reaction time methods important information about the time course of agreement processing remains opaque.…”

“…In Finnish, modifiers agree with their head nouns both in case and in number and the agreement is expressed by means of suffixes (e.g., vanha/ssa talo/ssa 'old/in house/in' → 'in the old house'). Vainio et al (2003;2008) showed processing benefits for this kind of modifier-head agreement, when the head nouns were relatively short. However, the effect showed up relatively late in the processing stream, such that word n + 1, the word following the target noun talo/ssa, was read faster when it was preceded by an agreeing modifier (vanha/ssa) than when no modifier was present.…”