From: LAURA SLOCUM In this article, acoustic and articulatory properties of /s/ and "th" were measured in order to ascertain if "th" for /s/ substitutions were distinguishable from target "th" productions. To this end, 5 children from each of 4 groups -- two "misarticulating" and two "normal" controls - participated in a word production task designed to elicit word-initial /s/ and "th" in 5 vowel contexts. Noteworthy is that the children assigned to the "misarticulating" condition had been in therapy for a period of 4 months. Individual and group differences in mean duration and relative amplitude of /s/ and "th" production were found. The authors concluded that "frontal misarticulation of /s/ cannot be characterized as a "th" for /s/ substitution" and that "even misarticulating children seem to have distinct phonological and phonetic representations for /s/ and "th" (p. 62). From: Kirk Baker This paper looks at temporal, intensity and spectral measures of fronted /s/ misproductions in children. Address two main points: 1) Is fronted /s/ realized as /T/; no. 2) Are there age related differences in misproductions; no. p.54 Why did they use an 8.5kHz low-pass filter but report Daniloff Wilcox Stephens 1980's /s/-fronters with energy peaks in 13-14kHz range (p.52)? Why didn't they use a sample rate high enough to look at this range of spectrum? p.58 Why are their fricative - vowel dB differences so much greater than DWS80? p.59 Is high centroid correlated with small front cavity? So centroid for /s/ should be higher than for /S/? If I don't have it backwards, I'm confused about why /s/ centroid 6369Hz is higher than /T/ 5730Hz. Is it because there's no cavity anterior to tongue-teeth constriction for /T/, and centroid reflects that anterior resonance? Kathleen Currie Hall Notes on Baum, Shari R., and James C. McNutt. 1990. An acoustic analysis of frontal misarticulation of /s/ in children. Journal of Phonetics, 18: 51-63. Summary This paper discusses the results of a production study of /s/ and /theta/ articulation for both normal children and those who produce fronted /s/. It is found that contrary to popular conception, fronted /s/ is not always /theta/; there are consistent differences between /s/ and /theta/ in productions of children with frontal misarticulation. There is evidence that children who have frontal misarticulation are simply developmentally delayed, as older children with misarticulation corresponded well with younger normal children. Details 1. Four groups: 5-6 year olds with normal /s/; 5-6 year olds with fronted /s/; 7-8 year olds with normal /s/; 7-8 year olds with fronted /s/ 2. Task: repeat stimuli of the form /sVt/ or /TVt/. 3. Measures: Duration. All groups had longer /s/ than /T/ on average. More consistent with normal children. Misarticulating children tended to have longer /T/ duration than expected -- so their /T/ sounds more /s/-like rather than vice versa. Movre variability among misarticulating children. 4. Measures: Amplitude. All groups had lower amplitude /T/ than /s/ on average. Variability was same across all groups. Perhaps this is the most salient distinction between /s/ and /T/ and so is most accurate among misarticulators? 5. Measures: Spectral frequency. All groups had higher centroid frequency for /s/ than /T/. Not true across all individuals -- perhaps some of the misarticulators have only a single representation for /s/ and /T/? Misarticulators tend to have lower frequencies for /s/ than do normal children; this is probably the source of the label "frontal misarticulation." 6. Older children were less variable than younger children. Older misarticulatos tended to behave like younger normal speakers, suggesting that frontal misarticulation is simply developmental delay. 7. "Frontal misarticulation of /s/" is clearly not a single factor. It is not clear, however, if it is due to fuzzy mental representations of /s/ vs. /T/ or articulatory inability. Questions 1. I wonder if perceptually all the "frontal misarticulators" sounded the same to the testers, or whether the measures on the screening test were taken objectively, with no sense of where the problem actually lies for each child. 2. Would lexical information play a role? For example, would "thought" vs. "sought" make children show more of a difference than "seat" vs. "theat"? Was there an effect of vowel quality? 3. Did the stimuli vary in terms of vowels only within each pair presented, or were /s/ and /T/ contrasted in presentation? 4. Why were tokens without release of the final stop excluded? How many did this involve? 5. What exactly is average intra-speaker standard deviation? Why is there only one value for each group? From: ejkong@ling.ohio-state.edu An acoustic analysis of frontal misarticulation of /s/ in children - Baum and McNutt This study investigated the acoustic property of frontal /s/ produced by misarticulating children, which previously is thought as the interdental fricative for /s/ substitution. The acoustic measurement of duration, amplitude and centroid of /theta/ and /s/ from two age groups of each misarticulating and normal children revealed that both misarticulating and normal children produce two fricatives as distinct sounds with each other. Two fricatives were consistently distinguished by thier amplitude differences across/within speakers, whereas they have more variability across/within misarticulating children than nomal children in durations and centroid values of two sounds. - In the results anlaysis, they put all the values from each 5 vowel together. Wouldn't it be meaningful at all to look at acoustic values of fricative before each vowel? - In general (for typically developing kids), what does the language developement stage of 5 year old and 7 year old kids look like? (How is it distinguished from the adult stage? Only the vocabulary size differences? ) From: Fangfang Li 1. (pp 55) they did not mention how they decided whether a given misarticluated /s/ is frontal or not. it'll helpful if they can include some info or definition of frontal misarticluated /s/. 2. (pp. 56) Table 1. i don't know what is exactly implied by inter-subject variability. but simply looking at Table 1, it seems that /th/ exibit bigger difference between normal speakers and misarticulating speakers for both age groups than /s/. it seems that they have the same problem of still not having enough subjects to make strong claim about it. 3. (pp 57) what is intra-subject variability and how is it calculated? what is the difference between intra-subject variability and inter-subject variability? 4. (pp 59) Table 3. it mentions in the following passage that "the centroid values computed for /s/ for both misarticulating groups were considerably lower than those for /s/ produced by each group's normally articluating peers, likely yielding the "fronted" percept." if we actually look at Table 3, it is true that the centroid values for /s/ target is lower for misarticulating groups, but the /th/ target has even lower centroid value (5538 for the misarticulating year group vs. 5877 for the normal group of 5-6 year old, and so on). it seems that the misarticluating groups still maintain the contrast between /s/ and /th/, although for both /s/ and /th/, but it is no clear as to how to interpretate the lower centroid values for /th/ in the disordered subjects. actually, centroid alone is a very gross description of the sepctral difference between /s/ and /th/, and it cannot capture the fact the two fricatives have very distinct spectral distrbution, with /s/ having a clear, well-defined peak in the high-frequency range, whereas /th/ very diffuse distribution of energy with no clear peak to be identified. From: nagar@ling.ohio-state.edu The study examines the way in which 5-6 and 7-8 year old normal an misarticulating children prnouned the fricatives /s/ and /T/ in different vowel enviornments. A comparision is made beteen the duration of /s/ and /T/ sounds. The study is aimed if the duration of these fricatives has any bearing on the way these fricatives are perceived and weather /T/ is being substituted for /s/ . The paper lacks in that it does not clearly explain the way in which the findings of the experiments conducted connected to the bigger picture. The targets for /s/ and /T/ are not explained clearly. I did not totally agree with the technique of elicting data. Could there be no circumstance where these two sounds occur in a (made-up) minimal pair? From: Helena Riha p. 52 "It would be interesting to determine whether frontal /s/ misarticulations truly appear to be /T/ for /s/ substitution errors." Are the authors also interesting in testing production of the dental affricate /tT/ (as in "eighth") and the alveolar affricate /ts/ (as in "cats") to see whether they are misarticulated? p. 53 "Is misarticulation a delayed normal development or a completely different course of development?" In light of reading last week's articles, it is interesting to consider how children's misarticulations of /T/ and /s/ can be explained in terms of their evolving phonological inventory. How correlated is the accuracy of their productions with their acquisition of a contrast between these two segments? p. 54 "A ... speaker presented two stimuli in succession as he pointed to the words ... written on index cards; e.g., 'This is a [sit]. This is a [sIt].'" The children were shown the words in orthographic form. Could reading the words have confused them and led to errors in production of the target segments? p. 62 "Older misarticulators may simply be functioning at a less mature stage in phonetic development." How accurately do these children hear and distinguish the target phonemes in the speech of others? Are they functioning at a less mature stage in perception as well as production? From: Junko Davis - This study shows that the misarticulating children do have separate representations for /s/ and /T/, despite that the frontal /s/ misarticulations are frequently considered to be more like /T/ than /s/. Questions: 1. "the disordered subjects are producing /T/ tokens with longer than expected durations (more /s/-like), rather than /s/ token with shorter durations" (p.60) -- Baum & McNutt offers a possible explanation for this, which can be refuted by their own finding based on the other acoustic parameter. Could there be another possible explanation for this finding? 2. What is implied by "the average intra-subject standard deviation" in p.57? It seems that both misarticulating groups have lower values than those of normal peers. 3. "two of the subjects in each misarticulating group failed to produce the spectral difference, suggesting perhaps that they do not have distinct representations for the two phonemes." (p.61) "one subject in each of the normally-articulation groups also failed to maintain the spectral distinction between /s/ and /T/, indicating that the contrast may be signaled by varying acoustic cues in normal production." (p.62) It seems to me that these two different assumptions are derived from the same fact that some subjects failed to produce the spectral difference, so these sound a bit like circular reasoning. Also, I am not sure whether I spotted these children in the TABLE III properly. There are two children in the 5-6 year normally articulating group whose centroid value for /T/ is actually higher than that for /s/, but I am assuming that they are not the ones. 4. "the centroid frequencies computed for the misarticulating subjects were considerably lower than those for normal subjects, consistent with the perception of inaccurate /s/ production." (p.62) -- What about the subject LH in the 5-6 year old misarticulaing group in TABLE III? How did the productions by this subject sound? Did they still sound as equally misarticulated as the productions by other children in the same group? From: ASIMINA SYRIKA -What falls within the boundaries of a good s? Is it dependent on the language/dialect of the individual listener or speaker? -What exactly is meant by functional misarticulation? -p.52 'Specifically, children's productions of a variety of speech sound contrasts, despite being perceptually adequate, are not comparable to the adult model until age 10 or 11...'. Perceptually adequate for whom? If it is just for the adults, why is the perception of the child 'sound contrast producer' or of other same-aged children not taken into account?