Blog

The Lupyan reading introduces the paradox of the universal triangle and shows through four experiments—drawing, speeded recognition, unspeeded visual judgements, and inference—that people’s representations of triangles are graded and context specific.  During each of these tasks, the kind of triangle evoked was not some universal pure abstraction, but rather the shape of the triangles varied depending on the task.  This illustrates how our thoughts about triangles tend to be about particular instances of them versus a platonic form of one.  This series of experiments also demonstrated how the type of language used to prompt each person affected the kinds of triangles produced during each task despite the fact that the words that prompted each participant were formally equivalent (triangle versus three-sided polygon).  Lupyan argues that this difference in representation speaks not to the universal abstractness of the triangle but rather to the human mind’s ability to modulate representations of different triangles.  The Lupyan article works against a classical bottom-up approach to the study of the mind where the mind is viewed as a symbol-processing device with “conceptualization…implemented via activation of a requisite symbol” (pg. 18).  Instead, this article sees the mind from a top-down approach where the mind is capable of storing mental representations about the external world (like different triangles) via some sort of read/write memory mechanism.

On an entirely different topic, the Rickford article talks about the unequal partnership that has historically existed between the academic field of sociolinguistics and the African American speech community.  Rickford argues that the American quantitative sociolinguistics have drawn heavily from this community including the development of variable rules and frameworks for the analysis of tense aspect markers, social class, style, and narratives.  Rickford argues that the academic community has not reciprocated that which it has taken from over the years—insufficient induction of African Americans into the sociolinguistics community, a lack of representation of African Americans in their writings, as well as the lack of involvement of African Americans in various roles in civil and academic service.  I think an equal partnership between these two communities is incredibly important from both an academic and social progress standpoint.  I feel as though the inclusion of rigorous academics from the communities being studied could lead to more enrichening and nuanced developments in their field.  There are certainly benefits in having a seemingly objective foreigner study a specific language—they can better describe the language objectively as an outsider looking in.  However, the inclusion of members in the targeted language community (in my case native Tagalog speakers in the study of Tagalog) can provide unique insights into the particular language that may be difficult to see for foreign observers.

A couple millennia short

The Lupyan article starts with a broad statement -- the way I've been taught not to start my own papers. What's funny about this particular one is that it really sells itself short. Modern academic linguistics might have started approaching the problem of abstraction three hundred years ago, but we have been thinking about this problem for much longer in the realm of philosophy. I'm really curious about whether there exist books in linguistics that directly interact with philosophy, especially classical philosophy.

One of the problems Plato sought to solve with his theory of Forms was precisely that of abstraction: how come we seem to have knowledge of universal concepts when we are only ever exposed to particular instances of these objects?

Aristotle then postulated that we possess a logical technique or capacity by which we subtract that which is incidental to an object and come to consider that object "qua triangle" or "qua dog" -- disregarding that which defines that particular triangle or dog. It seems to me that one of the things Lupyan is showing us is this process in reverse: starting with a concept and instantiating particulars. Though "triangles" and "three-sided polygons" are coextensive, they are dissimilar enough such that what comes of them is not exactly the same.

Though I am somewhat familiar with modern philosophy that complicates this in a myriad ways, I'm still drawn to the clarity of Aristotle and of Aquinas, who later brought together Aristotelian notions into a kind of empiricism that explains how we derive all knowledge from the senses. I have never read relevant academic work in the sciences that deals with these ancients directly. This might be because they are treated through the lens of the moderns, but still, I would like to see them addressed and brought into this debate.

Regarding the Rickford, I was pleasantly surprised to find a paper that both did proper linguistics and also a critical meta-analysis of one particular branch and the real-world impact it was having or not having. I was particularly struck by the idea of teaching children who learn to speak AAVE to read AAVE first before delving into SE. These are old ideas, too, proposed at the turn of the 1970s. The paper then convincingly argued for this pedagogic shift, and I became really intrigued by all the implications it would have if implemented again.

Blog 5

I must once again express immense regret at the extremely rushed pace of this class. This week, we're actually skimming over two huge topics at once: sociolinguistics (Rickford's examination of the relationship between the African American community and the academic sociolinguistic community) and semantics (Lupyan's argument that "triangle" doesn't necessarily refer to a mathematical, formal triangle in the minds of native English speakers).

Rickford's essay appears to be more "meta" than not, but in fact, he spends a great deal of the essay outlining recent findings about AAVE (African American Vernacular English), to support one of his claims that the Black community has given a lot to the study of linguistics.. While he only develops the ideas in those findings to a bare minimum, it actually provides a fairly good window into what sociolinguistics appear to do, as the studies he mentions span all of phonology, syntax, and morphology. Furthermore, Rickford's arguments here provide insight into some of the methods that sociolinguists use in gathering data, such as interviews and group sessions, and how their main conclusions are structured (for example, that consonant cluster simplification changes according to changes in social status gradually). At the same time, he criticizes some of those methods, showing how they may unconsciously promote racist views.
However, this is only one side of his argument. Rickford also states that linguists have not done much for the communities they study, especially considering that one of the primary motivations to study them was to help their economic advancement. This part of the paper hints at the status of linguistics in academia and frames a real problem, enhancing the introduction to sociolinguistics. (In fact, I would almost say it was a better choice than giving us a chapter from an intro textbook.)

Using Lupyan's article, on the other hand, has a much less "meta" approach - but for semantics, it seems to fit better. While the core focus on semantics, the meaning of "meaning", appears to be extremely philosophical in nature, the article demonstrates that although the hypotheses may stem from philosophical arguments, the studies can be conducted very objectively: Lupyan has subjects draw triangles and rate triangles based off of various linguistic stimuli, and includes an enormous amount of statistical calculations on that data to prove his point that concepts are not represented abstractly in the brain. Personally, while I don't doubt the data, the sample sizes seem relatively small (I admit, though, that it would be difficult to conduct the experiment on much larger sample sizes) and I'm slightly curious as to whether people who have studied advanced geometry would have a different concept of "triangle".

I'm nearing the 500-word limit, but there's one requirement I haven't really touched - integrating the conceptual information across readings. To be honest, this is pretty hard this week given how there are two different topics and how different the papers are from each other. Perhaps one can be applied to the other? Perhaps the selection of readings is to suggest that we focus on applying the fact that "abstract" thought isn't easy to primary education, in order to give back to our test subjects?

Socio- and Psycholinguistics

This week's readings seem to be an introduction into some more experimental fields in linguistics. In Rickford's paper, he discussed both the influence of AAVE on the study of sociolinguistics while also critiquing the linguistic community for not reciprocating the support towards the African American community. Then, with a shift into psycholinguistics, we got the Lupyan paper discussing how we can perceive something as simple as a triangle so differently if it is simply referred to using slightly different, but still equivalent language.

In regards to the Rickford reading, it reminded me of our first reading that showed how linguistics can provide insight into gender stereotypes in society. It reminded me how much knowledge about a society there is to gain simply by studying the language. However, as Rickford writes, that knowledge doesn't do anyone but academics a whole lot of good without actually giving back to the community that is studied. Sure we know more about AAVE and its structure, but how can that be put to use in inner city schools to help teach Standard English? It may have been experimentally proven that gender is performed, but how can that help us counteract harmful gender stereotypes? I feel like these sorts of real applications are the real benefits of the field of sociolinguistics. Learning more about language helps us learn more about people, but if we can't use what we've learned to help those people then really what is the point?

As for the more psychology-based Lupyan reading, I found it very interesting how we can perceive a visual representation of a basic shape differently just by using slightly different language. When thinking about this in relation to brain function, it is intriguing to think about why we make certain connections to particular words or phrases rather than others, leading us to draw a slightly different shape whether we hear "triangle" or "three-sided polygon".

In his paper The paradox of the universal triangle, Gary Lupyan uses simple triangle-drawing experiments to investigate the way our brains represent concepts and link them to words.  This question is an ongoing debate in science and philosophy, and one theory is that our brains store each concept as a single entity that can be evoked by any description of it.  For example, the concept of a triangle could be called up by ‘triangle,’ ‘three-sided polygon,’ or any other correct description.  If this were the case, then we would expect that asking people to draw triangles or three-sided polygons would yield the same results.  Lupyan performed this experiment, found that prompt wording actually made a big difference – when people were specifically asked to draw ‘triangles’, they drew much more standard figures (almost all subjects drew upwards-pointing equilateral triangles) than when asked to draw ‘three-sided polygons’.  This supports the idea that labels like ‘triangle’ evoke specific perceptual states.  According to this theory, called “simulation,” labels allow us to aggregate many different experiences with specific categories and thereby learn the common characteristics that define the category.  In the case of a triangle, most of these experiences involve equilateral, upward-pointing triangles, so when people access their internal representation of a triangle, that specific standard triangle gets recreated.  Asking people to draw a ‘three sided figure’ activates a different set of experiences and representations, leading to greater variability in the resulting figures.  In addition, Lupyan found that standard-looking triangles were actually recognized faster than atypical ones, reinforcing the idea that people are comparing visual input with an internal triangle representation that reflects the most common types of triangles. 

The connection between classification and language is one of the most interesting parts of Lupyan’s article.  He writes that “language impairments such as aphasia impair categorization in nonverbal contexts” (3).  I would not have expected language and classification to be so deeply related.  If language is necessary for categorization, then does this prevent animals from effectively classifying objects?  Could the need for categorization have driven development of human language? 

In Unequal partnership: Sociolinguistics and the African American speech community, John Rickford discusses how the study of AAVE has advanced linguistics and improved our understanding of the process of grammatical change.  Sadly, AAVE continues to be perceived as ‘less correct’ than SE, leading to discrimination in job interviews, court cases, and other situations.  Rickford discusses multiple ways that linguistics researchers and linguistics programs can help.  For example, he talks about the Industrial Language Training (ILT) program used in the UK, which worked with employers to help modify their negative attitudes towards people with different accents or dialects.  Another key area where linguistics research can be applied is childhood education.  Linguistics experiments have shown that reading instruction where AAVE-speaking students read the same stories in both AAVE and SE can be far more successful than standard lessons.  In one case, the resulting student score increase corresponded to 6.2 months of instruction in just four months.  However, these programs are often misunderstood by the public and face harsh criticism, so they are not widely used.  Hopefully linguistics research will continue to influence public education and lead to better learning outcomes.  

I appreciated reading about some of the ethical issues involved in linguistics, which Rickford raises --- specifically regarding the unequal partnership between the African American speech community and sociolinguistic researchers. It becomes very easy, and often unintentionally so, for academia to exploit the wealth of knowledge within a community (especially one that has been traditionally marginalized) without equal reciprocation for the knowledge and insights exchanged. Rickford discusses the concept of “service learning,” which he hopes will become more widespread in higher education through models at Stanford, Cornell, and other prominent universities. In my personal experience of service at Stanford, I have faced similar ethical issues and participated in discussions of related topics, especially in our Branner community service conversations. I think researchers, not simply those interested in community service and are involved with the Haas Center for Public Service, should be well-versed and cognizant of the principles of ethical and effective service, such as reciprocity and learning through partnership, respect for diversity, and humility. One can argue that research in academia is in itself a form of public service, as findings in research lead to a better understanding of the world, which in turn can aid us in making more informed decisions. Thus, research in all fields should be bound by those same principles in order to be ethical and effective, making best use of their findings rather than simply learning from a population and taking those insights with them. However, I would disagree with some of Rickford’s proposed solutions for bridging the observed unequal partnership. He describes activities such as “tutoring in math,” which, though not his ideal form of reciprocity, “are relevant, too, and [he] would rather see [them] do something than nothing, establishing firmly for [them]selves and [their] students the notion of ‘service in return.’” Such activities assume a level of superiority in offering blind returns for what is gained; a more ethical approach would take into account a community’s needs and wishes rather than flatly imposing whatever the taker deemed to be a sufficient exchange.

Lupyan, on the other hand, talks about how context affects a subject’s response --- a recurring theme we have seen so far in our study of linguistics. He cites a series of experiments which illuminates the difference in shapes drawn when subjects are asked to draw a “three-sided polygon” versus a “triangle.” This illustrates the gap in our minds, where concepts are stored on an abstract (and therefore idealized) level, and our physical environment, where concepts are instantiated in multiple (and oftentimes imperfect) ways.

Mental Representation of Words Across Cultures

In his paper, Rickford discusses the importance of the African-American Vernacular English (AAVE) to sociolinguistics. Though African-American speak has been noted in pop culture, I was fascinated to see a formalized study of AAVE, and shocked that very little study had been given to this subject previously. For example, Rickford lists several grammatical attributes of AAVE, such as removal of the present tense –s, use of “be done,” use of “finna,” use of “be,” etc. Though I’m sure many of us have recognized the use of “finna” for example, it’s interesting to see these patterns of speech being presented as a sociolinguistic study.

Lupyan’s paper also presents interesting thoughts about the intersection of psychology and language. He talks about mental representation of language, using a triangle as an example. He proves that there is a default perception of certain objects, such as triangles or dogs—so much that people can even determine a “better” triangle from a set of various scalene, isosceles, and rotated triangles. Lupyan also experimented with asking people to draw a “triangle” vs. a “three-sided polygon.” He found that people were more likely to draw a canonical triangle when asked to draw a triangle, whereas drawings were more variable for a three-sided polygon even though formally, a triangle is equivalent to a three-sided polygon.

Together, Rickford and Lupyan discussions made me wonder about the cultural influence in learning and representing language. For example, do the mental representations of words vary between someone who grew up learning regular English and someone who grew up closely acquainted with AAVE? Assuming that mental representations of words depend heavily on experience and familiarity, do mental representations differ for people learning English as a second language (aka people who’ve spent most of their lives in a different country/culture)? How does this effect one’s learning and comprehension abilities?

Blog 5

This week, the Lupyan reading explored the problem of abstraction using the concept of a triangle as a simple example. Though the simple definition of a triangle is a three-sided polygon, Lupyan tested how people mentally conceptualize triangles through several experiments that tested three claims: 1) a person’s mental representation is more specific than their abstract definition of a triangle, 2) representation can be altered by eliciting context (“triangle” vs. “three-sided polygon”), and 3) category label “triangle” elicits more consistent and prototypical representations than does “three-sided polygon.” Lupyan found that “triangle” and “three-sided polygon” evoked systematically different representations, a finding that shows people’s representations are not strictly definitional but rather prototypical.

The Rickford reading covered the topic of linguistic study of African American Vernacular English (AAVE), and the unequal partnership between the researchers and those researched. Quantitative sociolinguistics has gained much data from the African American speech community but has not done much to give back to the community. Rickford notes the insufficient representation of African American induction onto linguistics. He also feels that African Americans should have greater representation in writing, courts, workplaces, and schools, and teaching of reading and language arts. A problem with the lack of African American voices in the sociolinguistic, ethnographic, and folklorist writing is that the presentation of the speech community is very negative.

I found that inaccurate generalizations was a topic held in common between the two papers. African Americans are often stereotyped by people outside that group, and triangles are similarly stereotyped by people outside the study of mathematics. While people know that African American is a label for people with American citizenship and some amount of African ancestry, the representations brought to mind are often much more specific and involve characteristics that do not apply to all African Americans. In the case of triangles, though people know that triangles are simply three-sided polygons, the representations brought to mind are also more specific and less general. In both cases, the mental representations cover only a small subset of the possible instantiations of “triangle” or African American. More mathematical education may correct the conceptualization of triangles and more insights from African Americans and increasing their voice in society and academia may help to correct the conceptualization of African Americans. These papers relate to the real world and the sad truth that 15% adults cannot recognize skewed triangles as triangles and the unfortunate prevalence of racism.