Thesis (MA)--University of Stellenbosch, 2005. / Bibliography. Also available via the Internet.

Thesis (Ph.D.)--Hong Kong University of Science and Technology, 2005. / Includes bibliographical references (leaves 272-280). Also available in electronic version.

Thesis (Ph. D.)--Stanford University, 1984. / Abstracted in DAI-A 45/06, p. 1740, Dec 1984. "Vol. 1"--[t.p.] Includes bibliographical references (leaves 617-622).

Thesis (M.A.)--University of Canterbury, 2006. / Typescript (photocopy). Includes bibliographical references (p. 166-169).

Thesis (Ph. D.)--University of Texas at Austin, 2006. / Vita. Includes bibliographical references.

This thesis considers various questions about acceptability judgement tasks (AJTs). In Chapter 1, we compare the prevalent informal method of syntactic enquiry, researcher introspection, to formal judgement tasks. We randomly sample 200 sentences from Linguistic Inquiry and then compare the original author judgements to online AJT ratings. Sprouse et al., 2013, provided a similar comparison, but they limited their analysis to the comparison of sentence pairs and to extreme cases. We think a comparison at large, i.e. involving all items, is more sensible. We find only a moderate match between informal author judgements and formal online ratings and argue that the formal judgements are more reliable than the informal judgements. Further, the fact that many syntactic theories rely on questionable informal data calls the adequacy of those theories into question. In Chapter 2, we test whether ratings for constructions from spoken language and constructions from written language differ if presented as speech vs as text and if presented informally vs formally. We analyse the results with an LME model and find that neither mode of presentation nor formality are significant factors. Our results suggest that a speaker's grammatical intuition is fairly robust. In Chapter 3, we quantitatively compare regular AJT data to their Z-scores and ranked data. For our analysis, we test resampled data for significant differences in statistical power. We find that Z-scores and ranked data are more powerful than raw data across most common measurement methods. Chapter 4 examines issues surrounding a common similarity test, the TOST. It has long been unclear how to set its controlling parameter d. Based on data simulations, we outline a way to objectively set d. Further results suggest that our guidelines hold for any kind of data. The thesis concludes with an appendix on non-cooperative participants in AJTs.

This study sought to investigate how temporal relations are expressed in Mandarin and isiXhosa. The objective was to compare how two typologically distinct languages, one isolating and the other agglutinating, cope with the encoding of situations that occur at different positions on the time line. Data was drawn from questionnaire responses from, and interviews with, isiXhosa speakers who provided translations of a wide range of sentences from English into isiXhosa. The study revealed that isiXhosa relies on the grammatical category – tense – to encode temporal relations, while Mandarin relies more on lexical and pragmatic devices – the use of temporal adverbials and the implication of aspectual markings – to locate a situation on the time line. Typically, each sentence in isiXhosa must be marked for tense: temporal adverbials are optional elements and used to more precisely locate the situation on the time line. By contrast, in Mandarin, temporal adverbials have a more central function in that they independently express different positions on the time line: without temporal adverbials, it is extremely difficult to locate a situation on the time line in some sentences. Another important difference between the two languages was revealed in this study: isiXhosa grammar allows speakers to talk about situations in terms of their distance (past or future) from the speech time whereas Mandarin grammar allows its speakers to talk about situations in terms of the internal properties (e.g. completed, ongoing, etc.) of those situations. The study revealed that isiXhosa and Mandarin are similar in one important respect: both languages have no formal properties, i.e. overt tense markers (isiXhosa) and compulsory temporal adverbials or other temporal devices (Mandarin) for marking the temporal location of present situations. The study also revealed that both languages encode ‘pastness’ in terms of whether or not the effects of the situation in question still remain at speech time.

M.A. / The term locative refers to all words that have a locative meaning. They include locative nouns, i.e. nouns that belong to classes 16, 17 and 18. Locatively derived nouns, i.e. nouns to which the locative prefix and/or suffix or variants thereof have been added to convey a locative meaning, are also included. Word groups and certain nouns can also obtain a locative meaning by the addition of the locative particle or locative possessive particle. Various theories on the origin of the locative prefix, suffix and class prefixes are discussed. The word status of locatives is investigated, leading to the conclusion that locatives belong to the word class "noun". A morphological description of locative nouns as well as of the norm and variants of the locative prefix and suffix used with locatively derived nouns is provided. The morphological features of the locative particle and the locative possessive particle are discussed. As a result of the combination of locative morphemes with nouns in the derivation of locatives or particles that occur with word groups, certain sound changes occur. These changes can be organized logically and in detail by using phonological rules. In European languages the meaning of locatives is conveyed by adverbs and prepositions. The basic semantic feature of locatives is the feature [+ locality]. In certain instances it also has the feature of reference to an object. Locatives have a wide variety of possible syntactical uses, including the following: subject valence, object valence, descriptive (or adverbial) valence, complementary valence, qualificative and antecedental valence. Pronouns derived from the locative noun classes can serve as auxiliary verbs and conjunctions. In certain syntactical contexts the locative prefix can be omitted from locatively derived nouns

This thesis is a literary linguistic study of lexical innovation in fiction. It uses corpus linguistic methods and concepts of morphological theory to develop a new word typology and to examine new words as to their role in directing a reader's imagination and with regard to their frequency and distribution in classic English literature between 1750 and 1923. A 56 million word corpus consisting of a homogenous variety of texts converted from online literature databases serves as the basis for a chronologically structured new word extraction. This is carried out aided by the concordancer programme AntConc. The following three aspects are addressed in this research. The first attempts to explain why certain new words appear newer than other equally novel forms. It demonstrates that the factors influencing a word's novelty effect are wordlike-ness, morpheme content, and formal and semantic analogy. A new word typology is derived from these. A second main section focuses on stylistic aspects. If the words we use influence the way we think, as theorised in the principle of linguistic relativity, then forming new words and reading these should influence the way we think about what they describe. The second element identifies the strategies authors may use to affect their readers' associations through word formation. A third section is a frequency and distribution analysis of the new words extracted, taking historical developments, text mode and form, genre, and new word types into account. It adds quantitative data to the qualitative investigation preceding it, showing that verse and prose, text forms, and genres as well as time periods differ in the new words they produce and providing evidence for the characteristics of each.

Pillow & Flavell (1985) argue that the phrase 'look like' increases the tendency of young children to commit intellectual realism errors. The present study followed their procedures with the Block Arrays (which includes the Hidden Block task) and, in addition, included the identity task from a previous appearance-reality study by Flavell, Flavell & Green (1983). Forty-two three- and four-year-old preschoolers were presented with a variety of block arrays (Block Arrays task) and realistic-looking fake objects (Identity task) to observe. The subjects were tested on all of the stimulus items in one task before being tested on the second one. Half the subjects received the Block Arrays task first, half received the Identity task first. After the presentation of each array or object, the subject was asked a test question about its appearance while looking at it through a viewing tube. In the Look Like Condition, the test question included the words 'look like' and in the See Condition, the verb 'see' was used instead of 'look like'. If the subjects made any errors in the first condition (Look Like), they then received the second condition (See). No difference was found in the childrens' performance in the Hidden Block task but there was a difference in their performance in the Identity task. However, this difference cannot be attributed solely to the two wordings but rather to some interaction of task type and condition. Further analysis of the Look Like Condition revealed an Age-by-Task interaction in which the threes and fours performed differently in each task. A significant main effect for each of Task and Gender was also found in the Look Like Condition. / Education, Faculty of / Educational and Counselling Psychology, and Special Education (ECPS), Department of / Graduate