Some reflections on a friend's dissertation defense
This morning I attended my friend's dissertation defense. I don't think she'll mind me discussing it here, but I'll just refer to her as Susan. We met in 2008 when I started an MA in the Department of Second Language Studies (SLS), where Susan was already studying for a PhD. She's been a great, loyal friend and she's a brilliant linguist and researcher. On numerous occasions, I've called her for help with syntax or semantics concepts, and she always sets me straight. To top it off, she's one of my regular rock climbing buddies at the indoor gym here.
I wanted to write about the defense a little while the ideas are still fresh in my mind. I applied to and started graduate school somewhat hastily after finding myself living back in Indiana unexpectedly for some family affairs. Honestly, I don't think I was well prepared for the program I entered (or even for life as a grad student generally), and while I did well in courses, I often felt like I was struggling to understand a lot of the concepts and formalisms. It wasn't until my third or fourth year that I learned of something called "impostor syndrome", where a person feels like he's the only one who doesn't understand what's going on, and worries that others will discover that he's an "impostor". It turns out that a lot of graduate students feel this way. But now that I have more coursework and research experience under my belt, I often find that I come across a concept that stumped me when I first encountered it, and now it makes perfect sense. Ultimately, it's a chicken-and-egg problem--you're not really equipped with the knowledge and experience to grasp material in a new field, but you have to start somewhere.
From where I am now, I think I have a much better understanding of the kinds of research the SLS faculty do and how the various research is related. More importantly, I have a better grasp of the theoretical framework common to much of this research. Prof. Rex Sprouse in SLS, with his co-author Prof. Bonnie Schwartz (at University of Hawaii, I believe), proposed a framework called "Full Transfer/Full Access" (FTFA) in work in the mid-1990s, namely "L2 cognitive states and the full transfer/full access model" (1996). I had a hard time understanding this framework and the competing frameworks (Minimal Trees, Full Access (no transfer), etc.) when I first encountered it (in Dr. Sprouse's own class, which probably makes me especially pitiful...), but it makes a lot more sense to me now. FTFA seems to support / be supported by the research of several SLS faculty. At the very least, the theory seems compatible with much SLS work. FTFA, as I understand it, posits that the initial state of the L2 (that is, the state of the second language in the mind of a person just beginning to acquire it) is the L1 (the person's first language, as it exists in his/her mind) itself. The theory supports the notion of Universal Grammar (UG; basically the idea that we all have an innate capacity for language) and posits that this UG used to mentally process or parse language--whether it is for our L1 or L2 (or beyond). Full Transfer means that I fully transfer my L1 knowledge, and Full Access means that I have full access to UG (other arguments claim lesser access, perhaps due to a critical period in language acquisition). So if I use the same tool to process English (my L1) and Spanish (my L2), why is it so much slower and harder for me to process Spanish? Proponents of FTFA would argue that the difference lies in access to the mental lexicon, which, under this framework, embodies not just words and their meanings, but also their usage in the grammar. In an L2, this lexicon is less complete and we have slower access to it.
Susan's dissertation examines the processing of different kinds of verbs. She measured the reading times of L1 and L2 speakers of English for sentences with different kinds of verbs. Specifically, the participants read the first word on a screen, press a button when they're finished, then the word disappears and the next word appears. This allows her to pinpoint where people spend the most time in processing a sentence. She relied on something called a garden path effect, which is where we form one analysis of a sentence, but as we encounter a later part of the sentence, we have to abandon the initial analysis and reanalyze. A famous example is the sentence, The horse raced past the barn fell down. This is a passive sentence, i.e., The horse (that was) raced. The garden path effect here is that after we read barn, we think the horse ran past the barn. Then we get to fell down, and we don't know how to analyze it. Then, ideally, we realize that it's not the barn that fell down, it's the horse.
Susan focused on two types of verbs, called unaccusative and unergative verbs, and used sentences constructed so that a noun phrase is in the object position of the first verb (arrive or speak), regardless of whether these verbs normally take an object. An example might be As the baker sneezed the cook chopped vegetables. She found that reading times for L2 English speakers spike when they encounter the second noun phrase, presumably because they do not expect that verb (sneezed) to take an object (the cook), so they have to reanalyze--in fact, the second noun phrase is the subject of a second verb. Without going into detail about it... Because in one language, a verb may take an optional object, while in another language it cannot take an object, Susan was able to study groups from specific L1s, look at the properties of the verbs in that language versus English, and infer via the effect on reading times that the L2 English speakers do in fact seem to transfer their lexical knowledge from the L1 to the L2, which is a really neat finding. She also found that in comprehension questions about the sentences, L2 speakers are more likely than L1 speakers to retain the actual information from their incorrect initial analysis; i.e., Q: Did the barn fall down? A: Yes.
But here (in my opinion) is the kicker--the L1 English speakers exhibit a similar curve in reading times, it's just flattened out quite a bit. This lends some real support to the idea that we use a single system for processing an L1 and L2, with the differences in performance attributable to weaker access to the lexicon.
Congratulations to Susan on a wonderful defense (and apologies for my inadequate characterization of her research)! Now I'm off to celebrate with her and some friends. Happy Halloween!
I wanted to write about the defense a little while the ideas are still fresh in my mind. I applied to and started graduate school somewhat hastily after finding myself living back in Indiana unexpectedly for some family affairs. Honestly, I don't think I was well prepared for the program I entered (or even for life as a grad student generally), and while I did well in courses, I often felt like I was struggling to understand a lot of the concepts and formalisms. It wasn't until my third or fourth year that I learned of something called "impostor syndrome", where a person feels like he's the only one who doesn't understand what's going on, and worries that others will discover that he's an "impostor". It turns out that a lot of graduate students feel this way. But now that I have more coursework and research experience under my belt, I often find that I come across a concept that stumped me when I first encountered it, and now it makes perfect sense. Ultimately, it's a chicken-and-egg problem--you're not really equipped with the knowledge and experience to grasp material in a new field, but you have to start somewhere.
From where I am now, I think I have a much better understanding of the kinds of research the SLS faculty do and how the various research is related. More importantly, I have a better grasp of the theoretical framework common to much of this research. Prof. Rex Sprouse in SLS, with his co-author Prof. Bonnie Schwartz (at University of Hawaii, I believe), proposed a framework called "Full Transfer/Full Access" (FTFA) in work in the mid-1990s, namely "L2 cognitive states and the full transfer/full access model" (1996). I had a hard time understanding this framework and the competing frameworks (Minimal Trees, Full Access (no transfer), etc.) when I first encountered it (in Dr. Sprouse's own class, which probably makes me especially pitiful...), but it makes a lot more sense to me now. FTFA seems to support / be supported by the research of several SLS faculty. At the very least, the theory seems compatible with much SLS work. FTFA, as I understand it, posits that the initial state of the L2 (that is, the state of the second language in the mind of a person just beginning to acquire it) is the L1 (the person's first language, as it exists in his/her mind) itself. The theory supports the notion of Universal Grammar (UG; basically the idea that we all have an innate capacity for language) and posits that this UG used to mentally process or parse language--whether it is for our L1 or L2 (or beyond). Full Transfer means that I fully transfer my L1 knowledge, and Full Access means that I have full access to UG (other arguments claim lesser access, perhaps due to a critical period in language acquisition). So if I use the same tool to process English (my L1) and Spanish (my L2), why is it so much slower and harder for me to process Spanish? Proponents of FTFA would argue that the difference lies in access to the mental lexicon, which, under this framework, embodies not just words and their meanings, but also their usage in the grammar. In an L2, this lexicon is less complete and we have slower access to it.
Susan's dissertation examines the processing of different kinds of verbs. She measured the reading times of L1 and L2 speakers of English for sentences with different kinds of verbs. Specifically, the participants read the first word on a screen, press a button when they're finished, then the word disappears and the next word appears. This allows her to pinpoint where people spend the most time in processing a sentence. She relied on something called a garden path effect, which is where we form one analysis of a sentence, but as we encounter a later part of the sentence, we have to abandon the initial analysis and reanalyze. A famous example is the sentence, The horse raced past the barn fell down. This is a passive sentence, i.e., The horse (that was) raced. The garden path effect here is that after we read barn, we think the horse ran past the barn. Then we get to fell down, and we don't know how to analyze it. Then, ideally, we realize that it's not the barn that fell down, it's the horse.
Susan focused on two types of verbs, called unaccusative and unergative verbs, and used sentences constructed so that a noun phrase is in the object position of the first verb (arrive or speak), regardless of whether these verbs normally take an object. An example might be As the baker sneezed the cook chopped vegetables. She found that reading times for L2 English speakers spike when they encounter the second noun phrase, presumably because they do not expect that verb (sneezed) to take an object (the cook), so they have to reanalyze--in fact, the second noun phrase is the subject of a second verb. Without going into detail about it... Because in one language, a verb may take an optional object, while in another language it cannot take an object, Susan was able to study groups from specific L1s, look at the properties of the verbs in that language versus English, and infer via the effect on reading times that the L2 English speakers do in fact seem to transfer their lexical knowledge from the L1 to the L2, which is a really neat finding. She also found that in comprehension questions about the sentences, L2 speakers are more likely than L1 speakers to retain the actual information from their incorrect initial analysis; i.e., Q: Did the barn fall down? A: Yes.
But here (in my opinion) is the kicker--the L1 English speakers exhibit a similar curve in reading times, it's just flattened out quite a bit. This lends some real support to the idea that we use a single system for processing an L1 and L2, with the differences in performance attributable to weaker access to the lexicon.
Congratulations to Susan on a wonderful defense (and apologies for my inadequate characterization of her research)! Now I'm off to celebrate with her and some friends. Happy Halloween!
posted by leviaking at 3:27 PM
0 Comments:
Post a Comment
<< Home