... Finley & Badecker ... such an account cannot explain the gradience of learning data (ie, ... more ... Finley & Badecker ... such an account cannot explain the gradience of learning data (ie, non-categorical performance across participants and test items), and why height harmony learning data is less robust than back vowel harmony learning data (Linebaugh & Cole, 2005; Pycha ...
This paper presents a classification of vowel harmony, separating prototypical phonological harmo... more This paper presents a classification of vowel harmony, separating prototypical phonological harmony from morphemic harmony, which is triggered by the presence of particular morphemes. Morphemic harmony is driven by featural correspondence constraints requiring the realization of a feature across an output domain; phonological harmony is driven by markedness constraints requiring featural identity across an output domain. Morphemic harmony must always impose a fixed featural value on the form, which cannot be accounted for using approaches to phonological harmony. Phonological harmony cannot be incorporated into an analysis of morphemic harmony because morphemic harmony is represented only in terms of triggers, while exceptions to phonological harmony are generally represented in terms of targets. Typological predictions to the correspondence approach to morphemic harmony are also explored, including gapped representations, edge-only effects, and the variability of the realization of multiple features in morphemic harmony. Extensions of the present proposal are provided, illustrating that general featural affixation can be accounted for with morpheme-feature correspondence. Alternatives are also discussed. #
(1) It is possible to make sense of the intricacies of vowel harmony systems: triggers, targets, ... more (1) It is possible to make sense of the intricacies of vowel harmony systems: triggers, targets, dominance and directionality by adopting a Headed Feature Domains approach to harmony which applies a representation similar to that of autosegmental phonology, but applies ...
This paper argues that while the domain of regular vowel harmony processes applies over the entir... more This paper argues that while the domain of regular vowel harmony processes applies over the entire lexical item, exceptions to vowel harmony apply to a domain that is locally bound to the exceptional morpheme. This has important consequences for distinguishing between two competing theories of lexical exceptions in Optimality Theory (Prince and Smolensky, 1993Smolensky, /2004): lexically-indexed constraints (e.g., Pater, 2000) and lexically-indexed rankings (e.g., Anttila, 2002). Lexically-indexed constraints are subject to a locality requirement in their locus of violation, forcing exceptions in harmony to have a local domain of application. Lexically-indexed rankings cannot apply in a local fashion, and are unable to account for locality of exceptions in vowel harmony.
Poster presented at the Linguistic Society of America Annual Meeting, January, Chicago, 2008
-To ensure our results, we evaluated all possible inputs with all possible candidates using the C... more -To ensure our results, we evaluated all possible inputs with all possible candidates using the Contenders Algorithm (Riggle 2004) modified for the added level of representation. These results were fed into OTSoft (Hayes, Tesar and Zuraw 2003). - Contenders: - Create finite-state ...
The present study explores the effects of frequency in
learning to parse novel morphological patt... more The present study explores the effects of frequency in learning to parse novel morphological patterns. In two experiments, suffixes were divided into three classes: high, medium and low frequency, based on the proportion of stems in the input that each suffix attached to (high frequency = 12/12, medium frequency = 6/12, and low frequency = 2/12). In Experiment 1, learners were better at segmenting words containing high frequency suffixes compared to low frequency suffixes, even when the stems were novel. In Experiment 2, token frequency was controlled for across all three suffix frequency classes, but learners were still better at segmenting high frequency suffixes, even when words containing high frequency suffixes were less frequent. These results suggest that learners are sensitive to the frequency distributions of the morphemes in their language, supporting work suggesting that a Zipfian distribution may be ideal for language learning.
The present study explores learning phonological alternations that contain exceptions. Participan... more The present study explores learning phonological alternations that contain exceptions. Participants were exposed to a back/round vowel harmony pattern in which a regular suffix followed harmony, varying between /e/ and /o/ depending on the back/round phonetic features of the stem, and an exceptional suffix that was always /o/ regardless of the features of the stem vowel. Participants in Experiment 1 learned the behavior of both suffixes, but performance for the non-alternating suffix was higher when the suffix happened to adhere to vowel harmony. In Experiment 2, participants were exposed only to the same suffixes as Experiment 1, but the non-alternating suffix only appeared in harmonic contexts, creating ambiguity between exceptionality and alternation. Participants only correctly selected the non-alternating suffix when it appeared in a harmonic context. This suggests that learners are biased towards alternating harmony patterns, but require concrete evidence of non-alternation to learn the non-alternating suffix.
Nonadjacent dependencies are an important part of the structure of language. While the majority o... more Nonadjacent dependencies are an important part of the structure of language. While the majority of syntactic and phonological processes occur at a local domain, there are several processes that appear to apply at a distance, posing a challenge for theories of linguistic structure. This article addresses one of the most common nonadjacent phenomena in phonology: transparent vowels in vowel harmony. Vowel harmony occurs when adjacent vowels are required to share the same phonological feature value (e.g. V+F C V+F). However, transparent vowels create a second-order nonadjacent pattern because agreement between two vowels can ‘skip’ the transparent neutral vowel in addition to consonants (e.g. V+F C VT–F C V+F). Adults are shown to display initial learning biases against second-order nonadjacency in experiments that use an artificial grammar learning paradigm. Experiments 1–3 show that adult learners fail to learn the second-order long-distance dependency created by the transparent vowel (as compared to a control condition). In experiments 4–5, training in terms of overall exposure as well as the frequency of relevant transparent items was increased. With adequate exposure, learners reliably generalize to novel words containing transparent vowels. The experiments suggest that learners are sensitive to the structure of phonological representations, even when learning occurs at a relatively rapid pace.
Recent research has suggested a division between lexical representations and phonological pattern... more Recent research has suggested a division between lexical representations and phonological patterns; lexical items are stored with talker-specific information, while phonological patterns are represented at a separate, abstract level of representation . The present paper provides further evidence for this proposal, demonstrating that learners will extend a novel phonological pattern (vowel harmony) to speakers of a novel dialect when the words are familiar, but not when the words are unfamiliar, further supporting a distinction between the representations of lexical items and the representations of phonological patterns.
... Finley & Badecker ... such an account cannot explain the gradience of learning data (ie, ... more ... Finley & Badecker ... such an account cannot explain the gradience of learning data (ie, non-categorical performance across participants and test items), and why height harmony learning data is less robust than back vowel harmony learning data (Linebaugh & Cole, 2005; Pycha ...
This paper presents a classification of vowel harmony, separating prototypical phonological harmo... more This paper presents a classification of vowel harmony, separating prototypical phonological harmony from morphemic harmony, which is triggered by the presence of particular morphemes. Morphemic harmony is driven by featural correspondence constraints requiring the realization of a feature across an output domain; phonological harmony is driven by markedness constraints requiring featural identity across an output domain. Morphemic harmony must always impose a fixed featural value on the form, which cannot be accounted for using approaches to phonological harmony. Phonological harmony cannot be incorporated into an analysis of morphemic harmony because morphemic harmony is represented only in terms of triggers, while exceptions to phonological harmony are generally represented in terms of targets. Typological predictions to the correspondence approach to morphemic harmony are also explored, including gapped representations, edge-only effects, and the variability of the realization of multiple features in morphemic harmony. Extensions of the present proposal are provided, illustrating that general featural affixation can be accounted for with morpheme-feature correspondence. Alternatives are also discussed. #
(1) It is possible to make sense of the intricacies of vowel harmony systems: triggers, targets, ... more (1) It is possible to make sense of the intricacies of vowel harmony systems: triggers, targets, dominance and directionality by adopting a Headed Feature Domains approach to harmony which applies a representation similar to that of autosegmental phonology, but applies ...
This paper argues that while the domain of regular vowel harmony processes applies over the entir... more This paper argues that while the domain of regular vowel harmony processes applies over the entire lexical item, exceptions to vowel harmony apply to a domain that is locally bound to the exceptional morpheme. This has important consequences for distinguishing between two competing theories of lexical exceptions in Optimality Theory (Prince and Smolensky, 1993Smolensky, /2004): lexically-indexed constraints (e.g., Pater, 2000) and lexically-indexed rankings (e.g., Anttila, 2002). Lexically-indexed constraints are subject to a locality requirement in their locus of violation, forcing exceptions in harmony to have a local domain of application. Lexically-indexed rankings cannot apply in a local fashion, and are unable to account for locality of exceptions in vowel harmony.
Poster presented at the Linguistic Society of America Annual Meeting, January, Chicago, 2008
-To ensure our results, we evaluated all possible inputs with all possible candidates using the C... more -To ensure our results, we evaluated all possible inputs with all possible candidates using the Contenders Algorithm (Riggle 2004) modified for the added level of representation. These results were fed into OTSoft (Hayes, Tesar and Zuraw 2003). - Contenders: - Create finite-state ...
The present study explores the effects of frequency in
learning to parse novel morphological patt... more The present study explores the effects of frequency in learning to parse novel morphological patterns. In two experiments, suffixes were divided into three classes: high, medium and low frequency, based on the proportion of stems in the input that each suffix attached to (high frequency = 12/12, medium frequency = 6/12, and low frequency = 2/12). In Experiment 1, learners were better at segmenting words containing high frequency suffixes compared to low frequency suffixes, even when the stems were novel. In Experiment 2, token frequency was controlled for across all three suffix frequency classes, but learners were still better at segmenting high frequency suffixes, even when words containing high frequency suffixes were less frequent. These results suggest that learners are sensitive to the frequency distributions of the morphemes in their language, supporting work suggesting that a Zipfian distribution may be ideal for language learning.
The present study explores learning phonological alternations that contain exceptions. Participan... more The present study explores learning phonological alternations that contain exceptions. Participants were exposed to a back/round vowel harmony pattern in which a regular suffix followed harmony, varying between /e/ and /o/ depending on the back/round phonetic features of the stem, and an exceptional suffix that was always /o/ regardless of the features of the stem vowel. Participants in Experiment 1 learned the behavior of both suffixes, but performance for the non-alternating suffix was higher when the suffix happened to adhere to vowel harmony. In Experiment 2, participants were exposed only to the same suffixes as Experiment 1, but the non-alternating suffix only appeared in harmonic contexts, creating ambiguity between exceptionality and alternation. Participants only correctly selected the non-alternating suffix when it appeared in a harmonic context. This suggests that learners are biased towards alternating harmony patterns, but require concrete evidence of non-alternation to learn the non-alternating suffix.
Nonadjacent dependencies are an important part of the structure of language. While the majority o... more Nonadjacent dependencies are an important part of the structure of language. While the majority of syntactic and phonological processes occur at a local domain, there are several processes that appear to apply at a distance, posing a challenge for theories of linguistic structure. This article addresses one of the most common nonadjacent phenomena in phonology: transparent vowels in vowel harmony. Vowel harmony occurs when adjacent vowels are required to share the same phonological feature value (e.g. V+F C V+F). However, transparent vowels create a second-order nonadjacent pattern because agreement between two vowels can ‘skip’ the transparent neutral vowel in addition to consonants (e.g. V+F C VT–F C V+F). Adults are shown to display initial learning biases against second-order nonadjacency in experiments that use an artificial grammar learning paradigm. Experiments 1–3 show that adult learners fail to learn the second-order long-distance dependency created by the transparent vowel (as compared to a control condition). In experiments 4–5, training in terms of overall exposure as well as the frequency of relevant transparent items was increased. With adequate exposure, learners reliably generalize to novel words containing transparent vowels. The experiments suggest that learners are sensitive to the structure of phonological representations, even when learning occurs at a relatively rapid pace.
Recent research has suggested a division between lexical representations and phonological pattern... more Recent research has suggested a division between lexical representations and phonological patterns; lexical items are stored with talker-specific information, while phonological patterns are represented at a separate, abstract level of representation . The present paper provides further evidence for this proposal, demonstrating that learners will extend a novel phonological pattern (vowel harmony) to speakers of a novel dialect when the words are familiar, but not when the words are unfamiliar, further supporting a distinction between the representations of lexical items and the representations of phonological patterns.
Uploads
Papers by Sara Finley
learning to parse novel morphological patterns. In two
experiments, suffixes were divided into three classes: high,
medium and low frequency, based on the proportion of stems
in the input that each suffix attached to (high frequency =
12/12, medium frequency = 6/12, and low frequency = 2/12).
In Experiment 1, learners were better at segmenting words
containing high frequency suffixes compared to low
frequency suffixes, even when the stems were novel. In
Experiment 2, token frequency was controlled for across all
three suffix frequency classes, but learners were still better at
segmenting high frequency suffixes, even when words
containing high frequency suffixes were less frequent. These
results suggest that learners are sensitive to the frequency
distributions of the morphemes in their language, supporting
work suggesting that a Zipfian distribution may be ideal for
language learning.
learning to parse novel morphological patterns. In two
experiments, suffixes were divided into three classes: high,
medium and low frequency, based on the proportion of stems
in the input that each suffix attached to (high frequency =
12/12, medium frequency = 6/12, and low frequency = 2/12).
In Experiment 1, learners were better at segmenting words
containing high frequency suffixes compared to low
frequency suffixes, even when the stems were novel. In
Experiment 2, token frequency was controlled for across all
three suffix frequency classes, but learners were still better at
segmenting high frequency suffixes, even when words
containing high frequency suffixes were less frequent. These
results suggest that learners are sensitive to the frequency
distributions of the morphemes in their language, supporting
work suggesting that a Zipfian distribution may be ideal for
language learning.