Update prod.tex

chapters/prod.tex

@@ -355,18 +355,7 @@ \subsection{Methodology for acoustic phonetic analysis}\label{sec:phoneticmethod
 I aimed to conduct acoustic analysis on 30 tokens of \textit{like} for each girl. Whenever possible, these 30 tokens were made up of 10 tokens of quotative \textit{like}, 10 tokens of the discourse particle, and 10 grammatical tokens. The grammatical tokens were made up of a combination of lexical verbs and adverbs depending on what was present in the data.\footnote{Preliminary analysis provided evidence that the lexical verb and the adverb were phonetically similar to one another for a given speaker in terms of /k/ presence and the degree of diphthongisation.} This, however, proved impossible for all girls due to low token numbers, and fewer than 30 tokens were analysed for some girls. For girls with more than 10 tokens of a particular function of \textit{like}, the first 10 tokens extracted were analysed provided that they were unobscured by background noise. The token distribution is shown in \tabref{tab:tokensanalysed}.
 
 
-%summary(noapproxtype.lab)
- %        approxadv        dm        dp   lexverb      prep     quote 
-  %      0         0         0       292       104       100       239 
-%> summary(CRtype.lab)
- %         approxadv        dm        dp   lexverb      prep     quote 
-  %      0        22        58       160        48        49      119  
-%> summary(NCRtype.lab)
- %         approxadv        dm        dp   lexverb      prep     quote 
-  %      0        11        29       132        56        51       120 
-      %  summary(noapprox$CR)
-     %CR NCR 
-  %0 376 359 
+
 
 \begin{table}[t]
 \caption{The distribution of analysed tokens of \textit{like} for CR and NCR groups}
@@ -649,14 +638,6 @@ \subsubsection{Model 2: Quotative and discourse particle \textit{like}}
 
 
 
-%no luck with eps, why???
-%\begin{figure}
-%\epsfxsize0.9\textwidth
-%\epsffile{images/kCR.ps}
-%\caption{Graph of interaction based on model's predictions of likelihood of quotative \textit{like} depending on whether the /k/ is realised and whether the speaker is in a group who eats lunch in the Common Room (CR, blue) or not (NCR, red)}. \label{prodinteract}
-%\end{figure} 
-%\epsfbox{kCR.eps}  this doesn't work either...
-
 
 This interaction was not carried exclusively by quotative \textit{like}; the opposite trend of /k/ realisation was found for the discourse particle. While CR girls were more likely to produce the /k/ in discourse particle \textit{like}, NCR girls were more likely to drop the /k/.
 
@@ -666,7 +647,7 @@ \subsubsection{Model 2: Quotative and discourse particle \textit{like}}
 
 %I first observed this trend in the speech of Onya, who is a central member of the most rebellious group, The Real Teenagers. In order to find out whether other NCR girls also followed this trend, Onya's speech was not included in the model.
 
-%\subsection{Quality of \textipa{/ai/}}
+
 
 
 In regard to monophthongisation, mean pitch, /l/ to vowel duration ratio, and following environment, both discourse particle \textit{like} and grammatical functions of \textit{like} behaved similarly when compared to quotative \textit{like}. However, they involved different interactions: one with /k/ realisation and the speaker-specific probability of a token, and the other with /k/ realisation and the social grouping of the speaker. Whereas the realisation of /k/ was linked to frequency of use for quotative \textit{like}, it did not appear to be linked to frequency of use for the discourse particle. Phonetic differences between grammatical functions of \textit{like} and the discourse particle will be presented in the following section.
@@ -736,7 +717,6 @@ \subsubsection{Model 3: Grammatical and discourse particle \textit{like}}
 
 Also reaching significance in the model is an interaction between the /l/ to vowel duration ratio and whether or not the speaker ate lunch in the Common Room (p$<$0.05); tokens with a long /l/ relative to the vowel were more likely to be the discourse particle if produced by a common room girl and were more likely to be one of the grammatical functions if produced by a non-common room girl. Together with the interaction from model 2, the results suggest that a word's phonetic realisation can depend on a combination of the word's grammatical function and the social characteristics of the speaker who produced it.
 
-%  In NZE, the PRICE diphthong is un with a backer nucleus are considered more innovative  \citep{onzebook}.  This provides evidence that grammatical \textit{like} was more likely to be produced with a more innovative realisation than the discourse particle.  
 
 A summary of results from all three models is presented in \tabref{tab:sumprodresults}.
 
@@ -860,7 +840,7 @@ \subsection{Frequency effects}
 
 %The /k/ in quotative \textit{like} should be dropped for individuals who use a higher percentage of quotative \textit{like} compared to non-quotative functions of \textit{like}.  
 
-%[This suggests that the relationship between /k/ realisation and percentage of quotatives that were \textit{like} could be due to (a) social construction of identity through /k/ realisation or (b) predictability of producing quotative \textit{like} given the likelihood of that particular speaker producing a different quotative.  Both interpretations involve social motivation, either directly on the realisation of the phonetic variable or indirectly through the choice in which quotative to use.  ]this needs to be thought through more.
+%[This suggests that the relationship between /k/ realisation and percentage of quotatives that were \textit{like} could be due to (a) social construction of identity through /k/ realisation or (b) predictability of producing quotative \textit{like} given the likelihood of that particular speaker producing a different quotative.  Both interpretations involve social motivation, either directly on the realisation of the phonetic variable or indirectly through the choice in which quotative to use.  
 
 %...These two possible interpretations are not necessarily mutually exclusive; speakers express their identities through choosing which quotative to use and may realise the /k/ 
 
@@ -888,7 +868,7 @@ \subsection{Special status of discursive tokens}\label{sec:statusofdisc}
 
 
 
-%Work on grammaticalisation has shown a correlation between lexical frequency and whether a word has undergone an innovation in meaning.   Find: Traugott, E.C. 1989. On the rise of epistemic meanings in English: An example of subjectification in semantic change.  Language 65, 31-55.  And w
+%Work on grammaticalisation has shown a correlation between lexical frequency and whether a word has undergone an innovation in meaning
 
 
 
@@ -950,7 +930,7 @@ \subsection{Special status of discursive tokens}\label{sec:statusofdisc}
 Despite \textit{like} being associated with The PCs, girls in other groups were highly aware that they used it as well. Before the segment shown in (17), Marissa suggested that they try to avoid saying \textit{like} during the morning break. Ricky, knowing how frequently and automatically they used it, stated simply that ``it won't work''. The widespread use of the lexical items combined with their association with a particular group served to make the discursive functions of \textit{like} a target of socially-meaningful phonetic variation within the school.
 %(Interview 2:34)
 
-%An alternative interpretation is that quotative and discourse particle \textit{like} arecan occur with an identical preceding context, and to avoid ambiguity, speakers may maintain a phonetic contrast between the items.  Different speakers or groups of speakers could contrast them similarly to each other (as with monophthongisation) or differently (as with /k/ realisation).  Results show that not more of an effect for ambiguous contexts.
+%An alternative interpretation is that quotative and discourse particle \textit{like} can occur with an identical preceding context, and to avoid ambiguity, speakers may maintain a phonetic contrast between the items.  Different speakers or groups of speakers could contrast them similarly to each other (as with monophthongisation) or differently (as with /k/ realisation).  Results show that not more of an effect for ambiguous contexts.
 
 
 
@@ -1145,13 +1125,6 @@ \subsection{Storage of phonetic detail in the mind}
 
 
 
-
-
-%This interpretation is also consistent with (but not limited to) an exemplar-based model of speech production, where there is either a separate lemma level indexed to all of the function-appropriate exemplars or a ``lemma level'' that is not a separate stored exemplar but a result of generalizing over utterance-specific exemplars. 
-
-
-%-so girls at Selwyn High systematically produce different phonetic realisations of \textit{like} for its different functions and they do so differently depending on their social grouping, but is this information stored? Are the girls sensitive to these lemma-conditioned sociophonetic patterns? The following chapter will discuss results from a series of speech perception experiments conducted in order to shed light on this question.
-
 %\newpage
 %\thispagestyle{empty}
-%\mbox{}
+%\mbox{}