Feb 2 ===== Heffernan:2004 -------------- How exactly does one find out if effects within a person's speech actually come from their gender/ethnicity/etc or from the social factors that come along with those factors? (I.e. expectations for different people, how people are spoken to, etc.) I think it's difficult to work with data if even the method of measurement of said data is under discussion (chapter 2 mentions the best practices of measuring /s/). Have there been disagreements in the community because in different measuring methodologies? I found the descriptions in 2.2 and 2.3 very interesting. I'll have to think more about how I articulate in different languages. The author states that the way in which instances of /s/ are pronounced and the degree of breathiness during speech are used as a social marker by female speakers. However, the conducted experiment provides evidence that this effect is especially common among English speakers and not that much among the Japanese ones. In how far does the importance of the given marker depend on the language which we are considering? - p71: "Normal-sounding" adults sounds quite offensive to me and the authors don't state what this means. Is it not like female and male voices are on a spectrum, for example regarding pitch, where females generally have higher, male lower pitch, but there are males and females that have higher/lower pitch than expected? - p73: "(Linville) had naïve listeners identify sexual orientation based on speech samples, and reports an identification rate of 80%." As the identification rate is quite high, it seems like this experiment was valid. But I do wonder, if this kind of thinking might enhance existing stereotypes. Just as some people would assume by a person's appearance their sexual orientation, which is questionable, in my opinion. - If there are socially acquired components to language, would we expect to see change over a long period in time in this regard? Given that societies and the meaning of social gender evolve, the "extra meaning" indexed in the socially acquired component might change. - Couldn't there be other groups as well, that have adapted their own socially acquired component? The paper discusses the effect of breathiness for the construction of gender. I wonder if there has been research on how other phonation types (like creaky voice, harsh voice, etc.) may be influenced by gender expression? I was thinking about whether the results would hold for different speech settings/contexts, e. g. dialog with different people (maybe female vs. male dialog partner). How much would the performance of /s/ change, and in which direction? It is written, for the analysis, 25 tokens of /s/ (15 word-initial tokens, and 10 word-medial tokens) produced by 22 speakers (10 Japanese and 12 English) were used. I am wondering if it is a sufficient amount of data for such kind of research. Especially, taking into account SD (e.g. skewness for English female: mean - 0.148, SD - 0.457), it might be not enough. This paper was very interesting for me, since it worked as a kind of example for the paper Kate and I presented (Johnson 2005, Emergence of social identity). I was trying to do a quick search on similar papers for the German language, but didn't really get far yet. But I can imagine that the results would apply to German as well, or at least this would be an interesting topic to read more about. On p. 79, the author says: "If speakers and listeners are to use /s/ as cues to gender, they must be using the first moment (i.e. the frequency of the energy concentration) as their primary cue." How does that work? Which parameter of speech production does this correspond with, and how would speakers and listeners distinguish the other moments? Looking at Henton and Bladon's paper that associated breathiness with arousal, it seems like it was quite a speculative conclusion made from observing the experimental differences they found. And it's one of the primary sources this paper has for the social component of breathiness. The example of breathiness in Japanese also has nothing to do with gender, and more politeness. With the evidence they have here, I'm a little skeptical of the central hypothesis that HNR necessarily has a strong social gender component and not another factor. In the discussion section, Heffernan mentions the hypothesis that a small number of women/men produced the different fricatives due to their vocal tract anatomy, which causes the creation of stereotypical perceptions. I wonder how we can verify this hypothesis, since I don't think the actual data from the experiment can support much of it. In this case, I wonder if we can see the differences in articulatory production of the fricatives. I appreciated that the author surveyed a variety of languages when making the claim that /s/ displays markers of gender (though since gender is contextualized within a specific culture, it would be interesting to see if this was different for monolingual vs bilingual speakers - assuming the American English speakers were monolingual). With regard to the recording of participants' speech - I think it's interesting that they were made to read tokens in isolation, as opposed to at least reading a sentence, for example. This would make it obvious which sounds are being studied, no? 1. Since Hefferman chose breathiness as a mark of gender based on the findings of Henton and Bladon, and the results of his research support this assumption, can we conclude that "women intentionally use a breathy voice" consciously and on a consistent basis while speaking? 2. In the study, the partisipants' age ranged between 21 to 36. May we suppose that the results on gender marking would not change much among older participants? What would be interesting to see here, is not only the examination of /s/ with the breathiness, but also vowels like /a/ or /o/ and rhotization at the end of words. Although, I am not sure how quite usefull that would be for the categories age and social class, I think it would be interesting to see the results for social class and gender/ gender identity. The the examination of /s/ and breathiness do not show any results for Japanese participants. It would be fortunate, to find the result data, since Heffernan himself said, he could not find any results, or redo the experiment with the Japanese people, to get results and to compare them with these of the other participant groups that speak Spanish, Dutch or English. It also would have been interesting to see the report, what the acoustic correlation of ethnic identity are of the study with the articulation on two positions of the two variants of /s/. Comment: Where it seems that English has more focus on initial and medial /s/ sound changes, Japanese gendered differences were brought up, but I would like to emphasize how many there are and how unique they are to each other. You would generally assume that if someone is using "male only" words that they are either a male or want to be presented as males. For example, the "woman in male clothing" cliché is done very often in Japanese media (such as Noh and Kabuki theatre) and is more often way easier to tell when you know these gendered words (because almost all of the actors are male), when the character is finally revealed to be the other gender, they typically will change their entire speaking cadence. Question: It is brought up in the conclusion that anatomy plays to some degree a role in this assumption as does size of the person, do you think that these speculations hold true 18 years after the paper was released? With the better understanding of testosterone/estrogen therapy, trans voice therapy, and concept of gender role language I would think that many of these conclusions don't hold anymore.