Site Loader

How does semantic competition affect certain aphasic
patients and what can this tell us about lexical access?

During verbal communication, there can be occurrences in
which errors are made. These can arise from deficits in accessing lexical
information through semantics (Bormann, Kulke, Wallesch and Blanken, 2008). A
semantic error, or a paraphasia, is made when an individual selects a word from
their mental lexicon which although may be incorrect, bears a semantic
relationship to the target word (eg. apple -> pear).  A word that bears this semantic relationship
is known as a semantic competitor, which are words that become activated as one
searches their lexicon for the relevant material. Recent studies have shown a
relationship between the number of semantic errors an aphasic patient makes and
the number of competitors a target word has (Blanken, Dittmann and Wallesch,
2002).  Bormann et al. (2008),
investigated whether a large group of aphasic patients would present the same pattern
of results shown by an individual in Blanken et al.’s study (2002), in that the
amount of semantic errors increased when a target word possessed many semantic
competitors. Conversely, fewer competitors produced less semantic errors but
more errors of omission (cases of anomia). The results of their study on a
larger group of aphasic patients revealed equivalent results – a word with many
semantic competitors resulted in more paraphasias and fewer errors of omission,
whilst fewer semantic competitors held less semantic errors yet a higher number
of omission errors. Bormann et al. conducted this experiment whilst numerous
other variables were controlled for, such as word frequency and age of
acquisition. This study contradicted the predictions of the discrete two-step
model (Levelt, Reolofs and Meyer, 1999), one of the most detailed models of
lexical access, which claims that lexicalisation consists of two independent,
distinct stages. Despite having found that the error types under observation
must stem from overlapping processes and not from independent processing
sources, Bormann et al. could safely reject the ideas of Levelt et al., but
could not decide if their findings hinted at either an interactive model of
lexical access (Dell, 1986) or a cascaded model (Morsella and Miozo, 2002). However,
they succeeded in answering their titular question (Omissions and semantic
errors in aphasic naming: is there a link?) by discovering that both types of
errors must relate by the deficits causing these errors must both be found at
the word form level, and not at independent stages in a model of lexical access
as Levelt et al. (1999) claimed. Furthermore, they also provided evidence that
the size of a semantic category relates to the number or paraphasias made by an
aphasic patient. I agree that the size of a word’s semantic neighbourhood, and
therefore the number of semantic competitors it has, must influence the amount
of errors made by an aphasic when trying to retrieve said item from the mental
lexicon, due to the large amount of co-activated competitors that the patient
must then select from.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

A later study, conducted by Bormann
(2011), managed to replicate the results of the study of Bormann et al. (2008),
in the sense that this study displayed mirrored findings that semantic
neighbourhood size provoked semantic
errors – items with fewer competitors provoked fewer paraphasias, but target items with fewer competitors showed fewer
paraphasias but an increase in the number of omission errors. Bormann (2011)
suggested that this provides evidence that omission errors are affected by
semantic neighbourhood size, and therefore positions the deficit at the lemma
level in Levelt et al.’s (1999) model, contrasting the earlier findings that
omission errors occur at the word form level (Bormann et al., 2008). Britt,
Ferarra and Mirman (2016) investigated the lexical selection process during a
picture naming task, in which the presented item was either judged to have the
same meaning (alternate names condition)

or to have closely
related, appropriate meanings to the item (near semantic neighbour condition).
These responses were closely semantically related but not fully comparable to
the target item. Items of such had been labelled an ‘incorrect’ response in
older studies, so we can assume that they are with in the same error category
as paraphasias. Britt et al. (2016) examined the role of the left inferior
frontal gyrus in aphasic patients in part of their study, and found that
participants with aphasia had lowest accuracy when naming pictures in the near
semantic neighbours condition. Like the other studies mentioned, we can assume
that the level of inaccuracy these patients displayed was due to the fact that
the images had more semantic competitors, and therefore more errors were made
due to the more intricate choices needing to be made by the patient in
question.  They predicted that
individuals with LIFG damage should have particular difficulty naming low name
agreement pictures that require more difficult lexical and semantic selection
due to the amount of semantic competitors present. However, Britt et al. (2016)
found that the participants with left inferior frontal gyrus damage were no
more sensitive to those in the experiment without any damage, suggesting that
the number of semantic competitors in a picture naming task has little to do
with the accuracy of the target word being selected, contrasting their
predictions. Britt et al. (2016) stated the reasons and research provided by
prior studies to suggest that there is separate selection at lexical and semantic
levels or a single lexical-semantic selection step. Their study provided new
results that are consistent with selection by competition models (Levelt,
1999). Their experiments show evidence of distinct selection processes at
lexical and semantic levels, contrasting the claims of the experiment by
Bormann et al. (2008).

Mirman and Magnuson (2008) found
opposite effects of near and distant semantic neighbours: words with many near
neighbours (highly similar concepts, therefore having more semantic competitors)
were recognized more slowly than words with few near neighbours due to having
accessed a high number of semantic competitors, and words with many distant
neighbours were recognized more quickly than words with few distant neighbours
because there are less semantically related competitors to be accessed before
the target word is reached. From the data
produced from 62 aphasia patients, a higher rate of semantic errors for words
with many near semantic neighbours was revealed, and fewer semantic errors for
words with many distant semantic neighbours. This mirrors the results found by
Bormann et al. (2008) and Bormann (2011), in that the more closely related
competitors an item has, the more semantic errors are made. These results are
discussed considering attractor dynamics: a theory which is not often presented
in a linguistic setting. For an introduction to the basics of attractor
dynamics in cognition, see Spivey (2007), Mirman (2011). The study argues that
as near semantic neighbours are closely semantically related to the target
item, the influence of so many other competitors would slow the access to the
target word, indicating more errors. As distant semantic neighbours are further
from the target with fewer competitors, they do not create such a high degree
of semantic competition, and so the target item would be easier to access.
Mirman (2011) predicted that for aphasic patients, overall error rates are
likely to mainly reproduce the degree of impairment rather than the effects of
semantic neighbours because aphasic patients may have impairments at varied
levels of processing and produce different distributions of error types.
However, the study reported that the biggest effect of semantic neighbourhoods
was on proportions of semantic errors, indicating that semantic competitors do indeed influence the rate of errors made
in aphasic speech.

another less common exploration of lexical processing, Riley and Desai (2016)
compared the processing of concrete words to abstract words. They too stated the
evidence that semantic neighbourhood density has an impact on the accuracy of
word selection (Mirman and Magnuson, 2008). As neighbourhood density
relies on the number of competitors a concept has, it can be harder to compare
these two types of words. This is because concrete words (concrete nouns, for
example) have a more graspable meaning which can be accessed with more ease by
the mental lexicon through it’s related semantic competitors. Abstract words
are harder to relate to a specific concept, partly due to the smaller number of
semantic competitors it might have and furthermore due to the vaguer
associations matched with an abstract concept. Whilst their study did not
include aphasic patients, it is not difficult to see how aphasic processing can
be affected with this study in mind: if it is harder for an unimpaired mind to
access abstract terms, it can only increase in difficulty for someone with left
hemisphere damage as paraphasias and blocks (omissions) would further
contribute to difficulty in accessing a more indefinite concept. The results found by Riley and Desai (2017) suggest
that selection mechanisms operate at two distinct levels of processing
(semantic and lexical), rejecting the claims of Bormann et al. (2008)’s study
that lexical processing does not happen on two independent stages, but occurs
as an overlapping process.

(2012) found in a study that likewise to Bormann et al., that there is an
overlapping, or interactive process involved in lexical access. Yates states
that activation passes from one level to another, and then back, which
increases the activation for the target word and for it’s semantic neighbours.
The results from their studies support the feedback account of sematic
neighbourhood in lexical processing and that there is more activation
within the semantic system for words with many neighbours, and this activation
feeds back to the orthographic (written language) level. This further confirms
how the size of semantic neighbourhood enables lexical access. Although again
no aphasic patients were analysed in this study, it can be assumed that the
same assumptions hold – that the increase of activated semantic competitors in
turn increases the number of semantic errors an aphasic patient may make, due
to the large room for error caused by so many co-activated concepts.

size of a target word’s semantic neighbourhood, and therefore the number of
semantic competitors it possesses can affect an aphasic patient in the number
of semantic errors (paraphasias) that are made. If a word has a large semantic
neighbourhood and therefore more semantic competitors, there is a higher chance
for error as so many words are activated, it can be difficult for the aphasic
patient to select the correct one when so many similarly semantically related
items are also activated in their lexicon. Hence the findings of Bormann et al.
(2008) that more competitors a word has the more errors are made. Whilst
Bormann et al. attributed this to an overlapping process within a speech
processing model, stating that semantic selection errors occur at the word form
level, this is also backed up by other recent research, further rejecting the
idea of Levelt et al. that different errors operate on distinct levels of
processing. I too am unable to conclude the definite model of processing that
explains the occurrence of semantic errors as past studies cannot agree, hence
the addition of theories that are not often applied in linguistics.

I can assume that for lexical access in aphasic patients, deficits to the word
form level explicitly have a detrimental effect for both semantic errors and
errors of omission. It is worth noting that in the studies discussed, not all
explicitly involve aphasic patients. Some involved patients with anomia due to
Alzheimer’s or dementia (Bormann, 2011), but not specifically stroke aphasia.
As the symptoms and effects of stroke aphasia are closely related to those of
disease induced aphasia, we can assume that comparable results would be

Post Author: admin


I'm Erica!

Would you like to get a custom essay? How about receiving a customized one?

Check it out