Morphology and Issues with Meaning Flashcards
What are cranberry morphemes and what problem do they illustrate?
Cranberry morphemes are morphemes that end in the morpheme -berry. Some are composed of meaningful units, like ‘blackberry’ or ‘blueberry’, referring to color. However, some are composed of a non-sensical morpheme like ‘cran’ or ‘hickle’, which does not carry any meaning. These morphemes illustrate an interesting question in neurolinguistics. Do we decompose and store different morphemes? Do we always decompose? Do we only decompose when a word is made up of meaningful units?
Fill in
We can differentiate between different types of morphemes. Morphemes like -ed in worked are called ———— and ——————-. Some morphemes are not as ——————— but they are ——————-, like -hood while others can be ————————- and —————–, like -er in singer, villager or empty-netter.
We can differentiate between different types of morphemes. Morphemes like -ed in worked are called productive and predictable. Some morphemes are not as productive but they are predictable, like -hood while others can be unproductive and unpredictable-, like -er in singer, villager or empty-netter.
Fill in and answer.
Explain —————————-‘s Atomist Model and compare it to ————– and ——————-‘s Decompositional view.
Butterworth’s atomist model is a theory on lexical organisation. It states that each word is stored in their full inflected form. On the other hand, Taft and Forester’s decompositional view postulates that words are generated from morphemes, which are stored in the lexicon. As such, the two views are on opposite sides of the spectrum. The first one presupposes that we have an extraordinary amount of storage and that every single word is saved while the second one presupposes that we decompose everything and that we have an extraordinary processing power. Besides this presupposition, both have issues. The atomist view doesn’t answer how we create new words if all words are stored; it doesn’t show how words are related (stone - talk is the same as talk -talked). The second one, besides needing a huge processing load, if we truly decompose everything, the meaning of opaque compounds could mislead us (dead + line -> it isn’t really a line that’s dead, duh!)
Bad guy starts playing
In 2003, Longtin et al. conducted a research in which they compared reaction times of participants in a lexical decision task using the submilinal prime-target paradigm. They compared three conditions:
1. Transparent word (prime - gauffrette, target - gauffre)
2. Pseudo-derived (prime - baguette, target - bague)
3. Ortographic (prime: abricot; target: abri); ‘apricot’ vs. ‘shelter’.
They saw facilitation for the first two conditions and inhibition for the third.
Answer the following questions:
a) What was the target of this research?
b) Discuss the results
c) They also conducted a second research but this time, they used a cross-modal priming technique. In this version, there was facilitation only for the transparent condition and there was no effect on the other two. Discuss the findings of the second research and compare the two versions.
The target of this research was to assess the role of semantic transparency in French and in particular, to establish whether pseudo-derived words are processed in the same way as transparent words.
In experiment 1, because the paradigm is subliminal priming, there is no time to process anything but the form of the prime (semantics are not accessed). The results of facilitation for conditions 1 and 2 and inhibition for condition 3 show that the effects of priming are morphological and not ortographic. Our brains only decompose things that can be decomposable (masaža cannot be decomposed into masa + ža, makes no sense).
The second experiment used cross-modal priming in which the prime was an auditory one. Here, the effect is not subliminal and participants were sensitive to both the semantic and morphological overlap. Thus we can conclude that when we conciously perceive something, morphological relatedness is not enough and only semantically transparent words can prime base words significantly.
Briefly explain the prelexical affix stripping hypothesis (Taft and Foerster, 1975)
It states that words are decomposed into their morphological constituents even in the absence of semantic transparency.
Which of these is incorrect?
Homonyms have a
a) conceptual level
b) lemma level
c) semantic level
d) phonological leel
c) semantic level
Homonyms like bank [money] and bank [river] are an example of a ————— homonym, while pen [enclosure] and pen [writing] are an example of an ————— homonym. The difference is that…
Homonyms like bank [money] and bank [river] are an example of a balanced homonym, while pen [enclosure] and pen [writing] are an example of an unbalanced homonym. The difference is that they are either similarly frequent (balanced) or not (unbalanced).
Polsyemous words pose an interesting question in neurolinguistics - do we activate all possible meanings when we encounter a polysemous word or only a specific one based on context? How would you answer this question experimentally?
Well well well…If it isn’t polysemy. I would conduct a phoneme monitoring task, where I would compare reaction times of participants after an ambiguous word. A phoneme monitoring task asks participants to indicate whether they heard a certain phoneme or not. For example (push when you hear /b/)
The men started to drill before they were ordered to do so. -> polysemous
The men started to eat before they were ordered to do so.
Results of one such research indicate that participants take longer after a polysemous word and they speculated this is because we activate multiple meanings (when we encounter polysemous words).
In regards to the role of context on word processing, two main views evolved; the selective and the exhaustive access. First, explain the two views.
The second question is in relation to a phoneme monitoring study, which compared reaction times of two sentences:
a) The carpenter ordered the men to drill before they were ready. -> Longer RT!
b) The carpenter ordered the men to eat before they were ready.
Which view do these results support?
The selective access view states that context biases the interpretation and that only the intended meaning is accessed (if a word is contextually primed, there is facilitation of RT). The exhaustive view states that mutiple meanings are activated even when the item is in context. The results show support the exhaustive view, because even in context, reaction times were longer for the first sentences because participants had to activate multiple meanings. If they supported the selective acces view, there would be no (significant) difference between a) and b).
A research (Duffy, Morris& Rayner) examined lexical ambiguity and fixation times in reading. They focused on eye fixations on balanced and unbalanced homographs in a biased vs. an unbiased context. What they found was that with context, fixations on target word were longer only for unbalanced homographs. Without context, fixations on balanced ambigous words were longer than on control words. What do these results show? Focus on the different views of lexical access (selective vs. exhaustive)
Even when the context tells you to active the infrequent meaning (scale as type of skin), you activate everything , which points to the exhaustive meaning when homonyms are unbalanced. In balanced homonyms, you also have an exhaustive view (because you don’t know which meaning should be activated). You cannot say that only context or only balance is important; what is important is the interplay of the two (with context - unbalanced; no context - balanced).
I don’t really get these results but ??? whatever
The Cross-modal priming experiment (Swiney, 1979) examined lexical access during lexical decisions. It had three conditions: context (no vs. biasing); ambguity (yes or no) and prime-target contextuall appropriatness (appropriate, inappropriate, unrelated). An auditory prime (sentence ‘’The man [prime] was surprised when he found several spiders, roaches and other bugs in the corner of his room. ‘’) was followed by a visual target (ant, spy, sew; written).
Results:
facilitation only for: appropriate and inappropriate words with and without context in ambigous condition
(hear: bugs see: ant
hear: bugs see: spy
hear: spiders…bugs see: ant
hear: spiders…bugs see: spy)
and only appropriate words both with and without context (hear: insects see: ant
hear: spiders…insects see: ant).
What do these results show? Discuss what this tells us about the role of context in lexical access.
Both ‘ant’ and ‘spy’ were primed → show that BOTH meanings are activated even in a biased context.
What we can see is that independently of the context (related or inappropriate), we have facilitation of both. In an unambiguous example, only contextually related in biasing context the RT are facilitated. → in essence, how do we know we have facilitation after ‘bug’ for ‘ant’ and ‘spy’; by comparing mean reaction time of each of these conditions with the unrelated one (‘bug’ - ‘table’).
Lexical access is an autonomous process. Semantic facilitation was observed for lexical decisions to words related to both the contextually relevant and the contextually inappropriate meaning of the ambiguity, even in the presence of very strong prior semantic contexts. Semantic context does not direct lexical access. Rather, immediately following the occurrence of an ambiguous word all meanings for that word seem to be momentarily accessed during sentence comprehension.
Replication of Swiney (cross-modal, lexical access, hear bug see spy experiment) BUT with one small change. They showed the target three syllables after the ambigous (or control) word of a sentence. In this experiment, only contextually related words facilitated reaction times. What does this tell us about lexical access? Is this proof for bottom-up or a top-down effect
This shows that we activate everything and we select (and we need to make this decision otherwise we couldn’t talk to one another).
Lexical activation is autonomous and we can isolate it, at least when it has to do with the effect of semantically related context
It’s an indication of bottom up influence.
What is developmental dyslexia?
It has a neurobiological origin and often urns in families, but environmental factors might also contribute to it (40%). Dyslexia is independent of intelligence, education or socisocio-economic status. It can, however, result in significant educational and occupational disadvantages throughout the person’s lifespan. It affects 5- 15% of the world’s population, which equates to around 700 million people worldwide (International Dyslexia Association, 2017)
