Investigating activation changes as driving (L1-) attrition

Glodstaf, Walther

Permalink

https://hdl.handle.net/2142/130079 Copy

Description

Title

Investigating activation changes as driving (L1-) attrition

Author(s)

Glodstaf, Walther

Issue Date

2025-07-08

Director of Research (if dissertation) or Advisor (if thesis)

Montrul, Silvina Andrea

Doctoral Committee Chair(s)

Montrul, Silvina Andrea

Committee Member(s)

• Ionin, Tania
• Sprouse, Rex
• Talić, Aida

Department of Study

Linguistics

Discipline

Linguistics

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

L1-Attrition, Finnish, Priming, Activation

Abstract

This thesis tested if changes in activation could be registered in non-attriters for aspects of Finnsih grammar that had been previously documented to attrite. These grammatical areas were the accusative/partitive split, accusative case-form split, and pragmatic V2 word order. In so doing, this thesis attempted to adjudicate whether the explicit and implicit assumptions that L1-attrition is connected to changes in activation were empirically founded. It was also tested whether the size of the linguistic unit (morphology versus word order) mattered, and whether semantic meaning for morphology played a role, since the accusative/partitive split contributes meaning in the form of aspect and telicity, while the case-forms within the accusative do not. This thesis tested these grammatical properties of Finnish in a cross-linguistic priming experiment with a group of Finnish native speakers that had grown up in Finland and immigrated as adults to the US, Canada, Australia, and Estonia and lived there for at least 8 years. The split in the societal language (English versus Estonian) was chosen to see if results were sensitive to linguistic proximity of the second language. Results from a priming study indicated that morphology indeed could change activation levels, but only as a function of within-language priming, if the morphology in question was not shared 1:1 in terms of forms and syntactic features between languages. This speaks for the validity of assuming that changes in activation are behind L1-attrition. However, results were found primairly in the English-Finnish group and only for the morphological phenomena. This was attributed to the word order task not inducing priming due to the absence of sufficient context making the V2 word order pragmatically salient. Absence of direct priming results in Estonian-Finnish speakers was interpreted to possibly be modulated by Estonian-Finnish speakers being within native control norms in their Finnish usage across domains and interlocutors. There were cumulative primnig results for the accusative case-form though, which also supported the notion that morphology might not be able to prime cross-linguistically, due to forms and syntactic features between languages not matching 1:1. Only in exceptional cases when this match is present (such as with the accusative case-forms) would we expect cross-linguistic priming to be possible; which is what the cumulative priming effects signal. Finally, a longitudinal case study with an attriter using the same methodology as above is also looked into, whether attrition, once it has set in, induced a change in the grammar. This was done by analysing if primes containing attrited (but for the speaker acceptable) morphology would prime this same attrited output equally as well as the usage of this same morphology in a structure where it was grammatical. The speaker was first tested in a cross-linguistic priming task where they were identified as an attriter by producing ungrammatical unmarked accusatives when only a partitive or marked accusative was expected. Analysis of the results showed that this attrited production often happened after prior production of a grammatical form of the unmarked accusative that was possible to be produced in one of the fillers. To see if this possibility for grammatical unmarked accusative primed attrited production, the speaker was tested again 18 months later with the same stimuli and order of items, with the sole exception that this time the fillers where grammatical unmarked accusative was possible to be produced were changed to make this impossible. Results indicated that the speaker’s attrition was undergoing grammaticalization, as prior grammatical usage of the morpheme primed better than attrited usage. This suggests that for the speaker attrition was most likely a problem of inhibiting the activated morphology in contexts where it would be ungrammatical. This suggests that L1-attrition is also best understood as an issue of access rather than changed representations.

Graduation Semester

2025-08

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/130079

Copyright and License Information

Copyright 2025 Walther Glödstaf

Owning Collections