PSYCHOLOGY OF LANGUAGE

01 Jul PSYCHOLOGY OF LANGUAGE

Paper instructions:
!Discussion
At this point, you obviously dont have any data to interpret, but you should discuss what the possible outcomes of your

experiments are and how you would interpret them with regard to your question. What conclusions can you draw about the

cognitive theory in question given the various possible outcomes? Finally, if you have any thoughts about it, you could also

discuss future directions for this line of questioning.
!Appendix Stimuli !NOTES
!Deadline
The deadline is absolute. You will receive a full-grade penalty for each day that the paper is late (e.g., if you turn

your paper in on the 6th, your grade will be lowered by a full- grade).
!Quotations
Your paper should contain very few if any direct quotations. It is always better to reframe a passage in your own words

and cite it than to quote it directly. By doing so, you will make sure you fully understand what the authors are discussing.
!APA
Your paper should include inline citations whenever you use a published source as well as a bibliography at the end of

the paper listing your sources. The citations and bibliography should conform to APA conventions. A good overview of the APA

style may be found here: https://owl.english.purdue.edu/owl/resource/560/1/ (see links on the left side of the web page).
???!Sample Topics
!I have included 5 sample research topics to provide you a sense of the type of research
questions that would be appropriate for this assignment.
!1. Semantics?
Mirman & Magnuson (2008) demonstrated that semantic neighborhood density has different effects on processing depending on

whether the neighbors are similar or distant to the target concept. Processing is facilitated by having few close neighbors

and many distant neighbors. Although the authors propose that this phenomenon is best explained by an attractor landscape,

they do not consider other possibilities. The present work seeks to examine this issue by examining how other semantic

properties interact with similar/distant neighborhood size in order to better understand the phenomenon.?
?
Experiment 1: Does the abstractness/concreteness of a concept modulate the effect of neighborhood size? Abstract and

concrete concepts seem to be processed differently,
?
?with concrete concepts leading to faster reaction times than abstract concepts. This experiment examines whether the

effect of neighborhood density reported by Mirman & Magnuson is modulated by a concepts concreteness.?
?
Experiment 2: Does semantic class modulate the effect of neighborhood size? Hillis & Caramazza reported that concepts

from different classes (e.g., edible objects, living things, transportation, etc) appear to be represented in neurally

distinct regions of the brain. This experiment examines whether the neighborhood effect reported by Mirman & Magnuson obtains

for concepts from a variety of classes.?
2. Syntactic Processing & Bilingualism?
A variety of studies have demonstrated that syntactic comprehension is an interactive process, where listeners utilize

semantic, frequency, pragmatic, and discourse information in order to parse sentences. One area that does not appear to have

received much attention is whether bilinguals process sentences identically to monolinguals. This study investigates whether

information from a subjects native language (L1) affects the way they process sentences in their non-native language (L2).?
?
Experiment 1: Research has shown that syntactic processing is guided in part by the frequency with which a word appears

with a particular part of speech. For example, while backpack can be a noun or a verb, it is much more frequently

encountered as a noun. This experiment examines whether the frequency of a words L1 cognate affects how it is processed in

the L2.?
?
Experiment 2: Research has shown that syntactic structures can be primed: providing a subject with a particular sentence

structure (e.g., the passive voice) increases the likelihood that the subject will use that structure in their own speech.

This experiment investigates whether priming can be obtain across a bilinguals languages.?
!
3. Reading?
Theories of reading posit that there are two routes by which individuals may read a printed word aloud: a lexical route,

that retrieves the words phonological form from long-term memory, and a sub-lexical route, where the phonology is computed

on the basis of letter-sound correspondence rules. One area of communication that has not received much attention is texting.

While some texting abbreviations involve pronounceable words (e.g., LOL), others are identified by their letter names (e.g.,

BTW). Since this form of literacy is becoming more and more common, it is important to understand how individuals process

this kind of text. This study investigates whether subjects read various texting words using whole-word, lexical

representations or whether they convert words to letter names in order to access semantics.?
?
Experiment 1: The first experiment examines whether the phonology for pronounceable texting words are generated while

subjects read these words for meaning. That is, do subjects generate the phonological form “lol” when reading LOL, or do they

simply retrieve these words on the basis of their orthographic form??
?
!
?
?Experiment 2: This experiment investigates whether the phonology of a texting words letter names is generated during

written processing. That is, do subjects generate “bee tee double-yoo” when reading “BTW”, or do they simply retrieve these

words on the basis of their orthographic form??
4. Spoken Production?
Theories of spoken production propose that activation may cascade from one level of processing to the next. For example,

when producing the word CAT, semantic competitors such as DOG and RAT become active during processing and send activation to

their component phonemes in addition to CAT activating its phonemes. One question that has not yet been addressed is how far

information can cascade in this system. This study investigates whether information from semantics can cascade all the way to

articulation or whether there is a more limited flow of activation.?
?
Experiment 1: In general, words with concrete semantics (e.g., chair, butter, tree) tend to be processed more efficiently

and robustly than words with abstract semantics (e.g., addition, equality, altruism). This suggests that concrete words have

stronger, more robust semantic representations than abstract words. If this is true, concrete words may receive more

activation from semantics than abstract words, and if this difference can cascade to articulation, we may be able to detect a

difference using acoustic measurements. This experiment examines whether abstract and concrete words differ in their acoustic

properties.?
?
Experiment 2: In order to ensure that this effect has a semantic locus, the same measurements were conducted on words

that differ by their age of acquisition. Since age of acquisition also affects how well words are processed but is not

posited to have a semantic locus, the comparison between age of acquisition and concreteness will allow us to examine

different sorts of cascading effects in spoken production.?
?!
5. Spelling?
Research has suggested that the spelling process involves a working memory store, typically referred to as the Graphemic

Buffer. This store is hypothesized to maintain orthographic representations during the relatively slow process of spelling

(which occurs one letter at a time). One question of interest about working memory stores is their capacityhow many items

can they store at one time? This is of particular interest with the Graphemic Buffer given that languages differ in the

average length of their words. This study investigates whether the graphemic buffer has a fixed letter capacity or whether

the capacity can be expanded through exposure.?
?
Experiment 1: This study investigates the graphemic buffers capacity by comparing spellers of languages with two very

different average word lengths: English and German. Since German is an agglutinative language, its words are on average X

letters in length, which is significantly greater than the average English length of Y. If experience with German words

expands the capacity of the graphemic buffer, German spellers should show fewer errors on spelling long nonwords than English

spellers.?
!
?
Experiment 2: This study follows up on the first by determining whether it is possible to expand the capacity of the

graphemic buffer through experience. In this study, subjects were trained on progressively longer nonwords for 2 weeks. A

post-training test compared subjects who had been trained on long nonwords t