Dear students!
Michael Owren will be our next speaker on the topic "A proposed prehistory of language: Laughter evolution and early phonology" on the 12. of September - you can check him out under this website:
https://sites.google.com/site/michaeljowren/
The reading materials for his lecture are on the UMdrive (same website as before).
Chia-Cheng will present the paper: Owren Amoss & Rendall,
and Kenneth will present the paper: Owren et al in press
Presenters,
please let me know if you want to do a PowerPoint presentation - as we
learned the hard way, there is nothing set up beforehand in our room, so
I will need to do that! And, please, bring a memory stick and/or your
computer with the presentation.
Since the amount of reading
material will vary quite a bit over the course you might want to read
ahead, so you don't have such a big task every time the readings are a
little longer.
If there are still questions about the last two topics feel free to ask!
cheers
Uli
The article on laughter is particularly valid for my interests because like music, it often conveys feeling without language. In philosophy, music is often distinguished from the other arts because unlike painting, poetry, or theater; music does not seem to directly represent anything. Also, while music is like a language, it isn’t language. While we can interpret various phonemic and morphemic qualities, there isn’t a reciprocation process. However, music, like laughter seems to share qualities with the representational aspects of laughter in that it can illicit a response and allow for association based responses. Nevertheless, this seems to often happen automatically and in many ways seems to communicate many of the same things. If wonder if music shares similar signaling with laughter because without the sounds of laughter, the listener is less likely to illicit a positive response or response at all. As with every study, it brings about as many questions as it solves. For example, while the amplitude of the laugh is briefly discussed, to what degree the frequency or the modulation of the sound may have had an effect seems to be an important aspect not discussed. Would a laugh that has a particular cadence and pitch illicit a different response? Also, is there a correlation between types of laughs and interpreted meaning such as self-deprecation, nervousness, humor, etc… Lastly, are there types of voiced laughs that are less likely to be misinterpreted?
ReplyDeleteThe article “Organizing Principles…” seems to bring up a similar question that I asked in regard to last weeks articles: Novelty of sound. In this case the final pages of this article discussed novelty of sounds as a way to illustrate reception-first learning, dual-neural pathways, and vocal flexibility in primates. The reoccurring question for me is what would be the evolutionary purpose of novelty? While it shows flexibility, what does it do? Again, I think that it may have something to do with attention getting—not just in what it does, but also as a way to call attention to a sound to foster a learning environment.
This comment has been removed by the author.
DeleteIn Owren, Amoss, & Rendall.’s 2011 article, the distinction of “fixed signal” and “speech” was drawn according to their first proposed principle “production-first” versus “perception-first”. Fixed signals (e.g. cries and laughs) are proposed to be grounded in affect domain, whereas speech to be grounded in cognitive domain and thus more volitionally controlled. The authors later justified why they termed fixed signals production-first and speech perception-first vocal development, using evidence of auditory input being optional for producing cries and laughs but essential for speech production. In animal kingdom, only production-first vocal development is possible. This claim, however, seems to only emphasize the importance of speech perception in speech acquisition. In our previous seminar lecture, we know that other alternative ways are still possible for human’s development of speech. The question now turns to what fundamentally discriminate fixed signals from real speech, and this question, interestingly, is somewhat answered by Owren’s second article (in presss). Since laughter is a reliable signal that have evolved to carry positive content of the signaler, laughter (voiced but not unvoiced one) further evolves to elicit positive valence of the signal receiver. The different findings of voiced and unvoiced laughter are rather interesting too. In order to be selected for, a signal with characteristics that are stable and more auditorily discernable is preferred. Owren et al.’s experiments tested this reasoning and yielded results that support this statement.
ReplyDeleteOne curious question: how did Owren et al. ask their participants to rate emotion intensity without having individual differences of judging emotionality? I think emotional judgment is not as easy as it sounds. Some people would judge something “very positive” whereas others would judge “positive”. Does 7-point scale help balance participants’ judgment?
I found Owren, Amoss, and Rendall’s (2011) paper interesting is the idea of production-first and reception-first, which lead to the fundamental differences and similarities between human and non-human primates, as the authors indicate. For example, contrary to infant babbling, marmoset babbling-like vocalizations develop with certain level of complexity first and simplify them later. However, I wonder why the babbling-like vocalizations in marmoset need to be simplified later instead of showing optimal complexity as the marmoset group wants. Since these productions are (or presumably are) production-first development, the babblings should be useful at the beginning stage. If the babbling in marmoset community has no function, as the article suggests, why these sounds should be adjusted after its emergence?
ReplyDeleteThe second paper presents an idea that I had never thought about before. I do not know that people’s reactions would be different after hearing a laugh with and without voicing, despite the fact that laughs are fixed signals. From my first impression, hearing unvoiced laughs can imply something secrete is happening (e.g., laughing people behind their back). After reading this article, humans are really complex and simple creatures because we sometimes ignore subtle behaviors and try to focus on things that are very complex (e.g., guess other people’s thought). I like a sentence in conclusion: “human laughter is… instead primary function of influencing listener emotion and behavior”. Sometimes when I listened to a story, I will smile before the person finish a story because he or she smiled first, so I have a feeling that it will be a funny story. However, what the ratings of a fake unvoiced laugh or ironic laughs would be? Would listeners tell the difference between true and fake unvoiced laughs in the first place?
In Owren Amoss & Rendall (2011) they describe the difference between production first and reception first vocal systems. The authors point out that affect is very clearly involved when vervets are producing alarm calls. Owren et al (2011) explain that these vervets are visibly aroused and fearful when producing these alarm calls. They go on to explain that vervet infants display these types of vocalizations and that they are originally highly varied and over time they are whittled down and become more accurate. The infant vervets’ originally highly varied calls require no auditory experience for their production. Similarly, human infants are seen to produce spontaneous vocalizations (e.g., crying and laughing). Both spontaneous crying and laughing occur in profoundly deaf infants, and the acoustic variability between deaf infants and normally hearing peers has been shown to be insignificant.
ReplyDeleteI am very interested in the above phenomenon. It appears that through evolution nonhuman primates and human primates have an ingrained ability to express an emotional dichotomy, happiness versus sadness, or positive affect versus negative affect. I find this to be interesting because I can understand the evolutionary advantage of being able to express sadness or negative affect -- the ability for an infant to express that something is wrong (at least from their perspective), allows for caregivers to adjust what they are doing in order to respond appropriately to the infant vocalization. This makes perfect sense. What I do not understand, however, is the evolutionary advantage of being able to vocalize positive affect. Is not the absence of sadness enough for caregivers to know that the infant is well? Why then, is the vocalization of positive affect necessary?
Fredrickson & Branigan (2005) defend their “broaden-and-build” hypothesis in their paper “Positive emotions broaden the scope of attention and thought-action repertoires”. Their hypothesis explains that positive emotions broaden the scope of attention and thought-action repertoires. Thought-action repertoires refer to the set of actions and thoughts that are a result of some kind of environmental input. So, the environmental input of a predator being near would elicit a negative emotion to those who 1) see the predator and 2) those who hear alarm calls associated with the predator. This negative emotion will narrow the individual’s scope of attention and thought-action repertoire, generally meaning that the only available thoughts are now focused on fight-or-flight responses and/or having to do with survival. Positive emotions on the other hand are seen to broaden the scope of attention and thought-action repertoires.
My thought is, perhaps the evolutionary advantage of an infant’s ability to express/experience positive affect is that it allows the infant to have a broadened scope of attention and thought action repertoire. This broadening of their scope of attention allows them to explore their environment more efficiently. It would allow them to interact in more novel ways with their environment as well. Likewise, a key component in many newer machine learning algorithms is a random variable. It turns out, if the system allows itself to make sub-optimal decisions it can learn much more efficiently. I see this as analogous to infants’ positive affect. If infants only experienced negative affect and the absence of negative affect they would have a severely limited scope of attention, meaning they would interact with their environment in less novel ways, which would severely limit their learning.
This comment has been removed by the author.
DeleteI could also bring in some of the social psychology research on the primitive emotion contagion, which is defined as by Wild et. al (2001) as "the tendency to automatically mimic and synchronize movements, expressions, postures, and vocalizations with those of another person, and consequently, to converge emotionally."
DeleteSo, maybe the increased time a human infant spends under the care of their mother has been the evolutionary pressure for the infant to express positive affect nearly from birth. Maybe human mothers benefit from this positive affect expressed by the infant because, through the primitive emotion contagion, this positive affect of the infant is being spread to the mother. Which would then allow for the mother's thought action repertoires to be broadened.
I see that the authors of "Understanding Spontaneous Human Laughter: The Role of Voicing in Inducing Positive Emotion" describe infant expression of positive affect via laughter as a fitness signal, but I am arguing that this is not the only evolutionary advantage of the expression of positive affect.
DeleteEmotions are an embodied experience. For instance, happiness can be expressed through laughter and other physiological signals. The simple act of smiling, without emotional context has been seen to cause the person smiling to feel happy. An individual's assessment of their emotion (meta-emotion) is also an integral component of emotion. When an individual realizes that they are experiencing the physiological components of an emotion this can increase the valency of the emotion.
So, spontaneous laughter in infants does not necessarily have to be an expression of happiness, but instead a way to invoke happiness in the infant. From this perspective, you can argue that there is an evolutionary advantage for infants to feel happy.
After reading the Owren et al paper, I was reminded of a theory that is known in cognitive psychology as mood-state-dependent effects. Mood-state-dependent effects occur when material in a particular mood is recalled or recognized best when a person is tested under that same mood state. It is assumed that the mood at encoding will subsequently serve as an effective retrieval cue during recall. As such, I wonder how this theory can be tied to the work of Owren et al. Would learning to distinguish between voiced and unvoiced laughter in a particular mood result in the participants consistently rating voiced laughs as being positive and unvoiced laughs as being negative? For example, one experiment could have participants rate how positive or negative they felt in response to hearing voiced or unvoiced laughter after being induced into a particular mood state during encoding. I wonder if findings would be consistent with the current literature.
ReplyDeleteIn the first paper by Owren, "Two Organizing Principles of Vocal Production: Implications for Nonhuman and Human Primates," we again discuss "vocal flexibility," only this time instead of focusing on the extent to which vocal flexibility is unique to humans in many ways, we're looking at the extent to which it is also, actually, an important aspect of other nonhuman communication systems, as well.
ReplyDeleteThe two "organizing principles" as Owren calls them, differentiates between "production-first" and "reception first" in vocal development. To be honest, even after reading over the explanations for these two classification in the paper, I'm still a little confused as to what, exactly they're talking about. On page 531, production-first is defined as vocal development in which the "role of auditory experience or motor practice" is small and the role of "affective triggering of calls and limited flexibility in vocal acoustics" is more center-stage (for lack of a better analogy). Reception-first, contrarily, has a "central role for auditory and motor experience, cognitive control over vocalization, and flexible vocal acoustics" (531).
I understand the definition, but I think what confuses me is how the terms "production-first" and "reception-first" correlate to the given definitions.
On pp. 532-533, we're in so many words told that humans (if I have my definitions straight) are more reception-first and primates are production first. That is, "auditory experiences" are necessary in order for further vocal development to occur (such as the production of certain sounds and initiation of "babbling", etc. - p. 532-33). In order for human infants to produce sounds, these auditory experiences are not needed. However, if they are to continue down the path to human language, at some point, they will need the intervention of auditory experience to carry them onward.
So, I guess, now, after having written it out, it makes more sense. I think I get the difference, now, and why it's important.
Moving on. The second principle has to do with what Owren et al called "dual neural pathways" (531). That is, when a vocal system is developed in a language user (and I suppose in this case I'd be using language loosely here), it develops as a separate, parallel neural "pathway," rather than a "single multipurpose system" (531).
DeleteBeing somewhat of a stranger to the cognitive field, I think I still understand this to be discussing the process of expanding neural networks in the mind while learning and development takes place. It's fascinating how even the physical arrangement of these networks can tell us something about behaviors in which they function.
The example of the vervet monkeys was especially interesting. It is amazing the extent to which primates and other animals can communicate with one another. It makes me wonder though, what kind of studies have been done concerning cross-species communication. I'm sure that some are out there. I just don't know enough about the field to have been exposed to them, yet. But I'd certainly find those very interesting. And it seems, too, that such studies would give us, yet, another way to analyze the fundamental similarities (and differences) of animal communication systems by observing what each species has to rely upon in order to communicate with animals from a different species. Also, in the animal kingdom, how much does this even take place? I can imagine instances in which some animal either intentionally or accidentally crosses into some other specie's territory or a mother guarding her young. But how often and in what ways, apart from human-to-nonhumans, do other animals from different species communicate (non-violently)?
The second paper about laughing, "Understanding Spontaneous Human Laughter," explored some aspects to laughter I had never really considered before. That it proved to have an influential impact on other didn't surprise me, but what did was the difference in reaction that voiced and unvoiced laughter received. I'm not sure, exactly, what "unvoiced" laughter would be attributed to. I read the definitions on p. 6, but wished I could have heard an example. Continued reading of the experiments as I went along revealed a little more using "breathy" also to refer to unvoiced.
In response to one’s question about canonical babbling regarding its onset of emergence and the quality/quantity of babbling, here are some key findings collected from Oller (2000: http://www.amazon.com/Emergence-Speech-Capacity-Kimbrough-Oller/dp/0805826297):
ReplyDeleteCanonical babbling in special populations:
1) Socio-economic status (SES): the onset of canonical babbling and motoric development (e.g. rolling, reaching, sitting) does not delay sig. in infants from low SES (26 weeks for 13 LSES infants and 29 weeks for 15MSES infants). However, the total amount of vocalization in protophone categories of any kind, per unit time, did differ and it showed that MSES group to be the more voluble, with 7.5 utterances per minute compared with 6.0 for LSES group.
2) Prematurity does not seem to slow protophone development.
3) Bilingual or having different linguistic background other than English: canonical babbling ratios and vowel ratios were statistically indistinguishable across monolingual and bilingual groups.
4) Hearing experiences: the sooner deaf infants obtain hearing assistance, the sooner they may accumulate the necessary experience to trigger onset of canonical babbling (the correlation between age of onset of canonical babbling and age of amplification in deaf infants was very high =.69).
5) Total lack of hearing: found in cochlear aplasia, in which the hearing organ does not develop. Lynch et al (1989) studied one case like this systematically and provided clinical treatment associated with tactual vocoders. Tactual vocoders are devices that receive sound through a microphone, divide the sound into a number of channels based on acoustic frequency, and then use the information in each channel to drive a stimulator on the skin. A few months after the child began to produce canonical vocalizations consistently.
6) Middle ear problem such as otitis media: children with otitis media studied by Oller and his colleagues did no find a delay in their onset of canonical babbling.
7) Down syndrome: both Dodd (1972) and Smith & Oller (1981) found a great amount of similarities both qualitatively and quantitatively between children with down syndrome and children with normal development. However, a slight delay in onset of canonical babbling was found later researches.
The importance of canonical babbling:
1) The emergence of canonical babbling and protophone development is highly canalized even in special populations.
2) Multiple channels other than auditory input appear to be accessible by the acquisition of canonical babbling.
3) can serve as a clinical indicator and red-flag an emerging disorder long before traditional symptoms can be identified (as in Oller et al.’s 2010 PNAS paper: http://www.pnas.org/content/107/30/13354.full).