Word Meaning Research Paper

1.    Approaches To Word Meaning

When people think  casually about  the topic of word meaning,  they  typically  think  of  dictionaries.   They view dictionary entries as constituting a statement of a word’s meaning. They also think  of the dictionary  as the ultimate  authority on the correct meaning: in any dispute, they defer to the dictionary definition as determining  what  is right.  In  doing  so,  they  adopt implicit assumptions that word meanings belong to a language,  as opposed  to its users, and  that  there are experts who alone are qualified to report the true meaning of each word.

Elements   of  this  sort   of  view  appear   in  some linguistic and philosophical approaches to word meaning. For instance, language is sometimes treated as an abstract system to be analyzed, without reference to the users of the system. And Putnam’s (1975, Chap. 12)  influential   philosophical approach  to  meaning appeals  to a ‘division of linguistic labor,’  suggesting that correct use of some common words can be determined  only by experts who have special knowledge of the entities in question.

However, modern lexicographers—those who write dictionaries—do not  hold  such  a  view  themselves. Their job in providing  dictionary  definitions  is not to report  the ‘true’ meaning  of a word  that  is divorced from what ordinary speakers of the language use the word  to mean.  The meanings  associated  with words change over time, as their use by speakers of the language  shifts.  For  instance,  ‘nice’ currently  has  a highly  positive  connotation but  in  earlier  times  referred   to  someone   who  was  ignorant   or  foolish; ‘hussy’   meant    a   housewife,   and   ‘mistress’   was simply   a  feminized   form   of  ‘mister’  (Hock   and Joseph   1996,   Chap.   7).   Just   a   generation    ago, ‘mouse’ referred  only  to  a  small  rodent;  now  it  is used more often to refer to a computer peripheral. The goal of lexicographers  must  therefore  be to  capture something about  how a community  of speakers uses a word at a particular time in the history of the language (see, e.g., Laird 1978).

Psychologists interested in word meaning take a perspective more similar to that of lexicographers than to  either  the  layperson’s  intuitive  view or  the  view reflected in the kinds  of linguistic and  philosophical approaches mentioned. Psychologists  view meanings as  changeable,  as  belonging  to  individual  language users, and as shared  across communities  of language users.  But  psychologists,  unlike  lexicographers, are particularly interested in mental representation and in the cognitive processes that draw on mental representation.  Their   primary   goals  in  the  study   of meaning are to characterize the knowledge that constitutes  word  meanings  and  to describe how this knowledge is used in interpreting words heard or read.

2.    Meaning As Concepts

A standard way of describing what constitutes  a word meaning,  for a psychologist,  is to say that  meanings consist of concepts. Concepts are coherent nonlinguistic  knowledge   about   some  entity  or  set  of entities (e.g., a person could have a concept of his or her brother  Max and also a concept of what brothers are  in general).  A strict  one-to-one correspondence does not exist between the words known by a person and  his or  her  concepts,  since a person  can  have  a concept without having a word to label it. (Children in the early stages of language  learning presumably  are often  in this  situation.) Also,  a person  occasionally may know a word but have little or no meaning associated  with that word. In most cases, though,  the words stored in a person’s memory will have meanings associated  with  them,  and  the  argument is that  the stored  meaning  for each word consists of a concept. Thus, for any given person,  the meaning of ‘brother’ consists of his or her concept of what a brother  is, the meaning  of  ‘run’ consists  of  his  or  her  concept  of what running is, and so on. For many common words, the  associated  concept  is assumed  to  be similar  for most people, and so the meaning is shared.

The idea that word meanings are concepts, though, does not delve very far into the question of the nature of meaning.  If meanings  are concepts,  and  psychologists want to understand the nature of meanings, they need to know in more detail what concepts consist of. Psychologists have devoted much effort to the study of concepts, both for their own sake and in order to understand word meaning (e.g., see Natural Concepts, Psychology of ).

An early answer to the question  of what  concepts are like was suggested by linguists pursuing  their own interests in word meaning. The proposal  from linguistics was that the meaning of a word is a set of defining features;  that  is,  features  true  of  all  things  appropriately  called by that  name and  together  separating those things from all things called by other names. For instance,  defining  features  for  the  word  ‘bachelor’ might   be   ‘adult,’   ‘male,’  and   ‘unmarried.’   This proposal  was closely related  to  analyses  by anthropologists  suggesting  that  kin terms  such as ‘mother’ and ‘grandmother’ could be analyzed in terms of elements  of  meaning  such  as  such  as  sex (male  vs. female)  and  generation   (one  removed  vs.  two  removed). Psychologists initially adopted the defining features view as a view of the nature of concepts, not of word meaning. Under the perspective that word meanings consist of concepts, though,  the circle back to the original linguistic theorizing is closed, with the difference that  psychologists  considered  defining features to constitute  the mental representation of the meaning,  not  a  meaning  detached   from  individual minds.

By the mid 1970s, psychologists  were finding  this analysis  of  concepts  and  meanings  problematic  because for many words, it was impossible to determine defining features. Rosch and Mervis (1975) suggested that concepts consist, instead, of sets of features having varying strengths  of association  to the category.  For instance, most fruits are juicy but a few are not; many fruits are sweet, but some are not; some have a single pit while others have many seeds. The most common features, like sweet and juicy for fruit, are true of prototypical examples but are not defining features. If word meanings are represented by concepts, then according to this analysis, they are represented by concepts that include prototypes and probabilistically associated  features, but not defining features.

Refinements  and modifications to this view of concepts  have  been made.  Most  notable  is the  idea that  concepts are not simply unstructured collections of  features;   instead,   they  are  held  together   by  a ‘conceptual   glue’  that   provides  information about the  relation  of  features  to  one  another   and  to  the entities in question  (e.g., Murphy  and  Medin  1985). For instance, a person’s concept of birds might include the information that they all have wings, that most fly, and that some nest in trees, but these are not unrelated pieces of information. Instead,  the  concept  includes the  knowledge  that  having  wings  has  a  particular causal  relation  to flying and  being able to fly has a particular causal relation to being able to nest in trees. If concepts include such information, then word meanings, likewise, do.

3.    Multiple Meanings: Polysemy

Research  on  concepts  has  thus  by now  provided  a fairly  detailed  (though  still  evolving)  view of  what concepts consist of. In that sense, word meanings are also understood in some detail. However, saying that meanings  are concepts  leaves some important  issues unresolved.  These issues revolve around the fact that many words have more than one meaning. In a small number of such cases, a word has two unrelated meanings,  such as ‘ear’ meaning either a sense organ or a part  of a corn  plant.  In  these cases, it is often historical accident that the two meanings are expressed through the same word. During  the earlier history  of the language, two more distinctly different words may have been used, but they evolved into the same form as changes  to  the  sound   or  spelling  patterns   of  the language took place. Of greater interest for theories of meaning  are cases of polysemy, where a single word has two or more  meanings  that  are related  in some way. For  instance,  the word  ‘run’ means  something somewhat  different, but not entirely so, in each of the following  phrases:  ‘run  a  race,’  ‘run  errands,’  ‘run for   election,’   ‘run   a   company,’   ‘the   machine   is running,’   ‘the  water   is  running,’   ‘the  movie  runs three hours.’ In other cases of polysemy, a word may be used  in more  radically  different  (yet still clearly related)  ways.  For  instance,  the  word  ‘newspaper’ most typically refers to a physical object consisting of printed  pages of newsprint,  but in sentences like ‘The newspaper   is  being  sold’  or  ‘The  newspaper   says Jones is leading in the polls,’ the same word is being used to refer to the corporate entity that produces the physical object, or to the writers whose writing appears in the physical object. Sometimes a word is even used as a part  of speech it has not been used as before, as ‘siren’ is in  ‘The  police  sirened  the  car  to  a  stop’ (Clark  and Clark  1979). The challenge for a psychological  view of  meaning  is  to  explain  how  mental representations accommodate the  different  senses a word may have, and how the intended sense is understood.

One approach to addressing this challenge is to say that  people  have  a  single  meaning  stored  in  their memory for a given word, and variations  on that meaning  are  created  by the  context  it occurs  in.  A person  would  take  into  account   his  or  her  stored concept for the word along with other information in the sentence and  discourse  context  to determine  the meaning.  In some cases, there may be regularities  in what kinds of meanings can be constructed that allow people to quickly arrive at the relevant meaning. For instance,  there  may be a general  rule that  says that names of objects can be used to stand for people associated with those objects, making figuring out the meaning  of cases like ‘The newspapers  says Jones is leading in the polls’ straightforward (Nunberg  1979). In  other  cases, there  may  be no  standard rule,  but other  aspects of the context,  including knowledge  of the background and goals of the person  who said or wrote  the  sentence,  are  sufficient  for  the  reader  or hearer to figure out the right meaning.

This view assumes  that  the  word’s meaning  itself changes in the contexts. An alternative  way of framing what happens is to say that interpretation of the word is modified by the context, but the word meaning itself stays the same. For  instance,  the meaning  of ‘newspaper’  is  always  about   a  physical  object  made  of newsprint,  but  a person  can draw  on that  meaning, along with any standard rules and various clues from the sentence and discourse context,  to figure out that the speaker  is using the word ‘newspaper’ to refer to something somewhat  different. This alternative  framing  is more  consistent  with  the  idea  that  meaning consists of concepts, because it distinguishes  between meaning  as  stored  knowledge  and  the  use  of  that stored knowledge to construct  an interpretation on an occasion of use.

Whichever  framing  is used,  it is unlikely  that  all variations  in interpretation / meaning  are constructed anew in interpreting a word  in a particular context. One  reason   for   thinking   so  is  that   it  has   been impossible to determine  ‘core’ meanings  from which all other meanings can be derived. For instance, there is no set of features associated with the central meaning of  ‘run’  that  seem  to  be  true  of  all  the  senses  of ‘run.’  In  some  cases,  senses  of  a  word   seem  to develop in chains (Lakoff  1987, Case Study 2), where one sense derives from another, and another from that one, until there are senses having little or nothing  in common   with  the  typical  sense.  For   instance,   the concept   associated   with  ‘box’  in  its  most  central sense is something  like a rectangular container  made of cardboard for storing solid objects. However, there have come to be cardboard juice boxes (for drinking, not  storing,  and  for liquids,  not  solids). From  those developed   plastic   (not   cardboard)  reuseable   juice boxes, and from them, plastic juice boxes in the shape of various  animals  (not rectangular). A plastic bearshaped juice box has none of the properties  associated with typical  boxes,  yet people  willingly call it ‘juice box.’ Thus it seems likely that  people learn and store senses for ‘box’ that are distinct from and in addition to their stored  knowledge  about  ordinary cardboard boxes. Another  reason for thinking that at least some of the variations are stored in memory is that some are encountered repeatedly (as in the case of ‘juice box’), so it would be inefficient to have to derive them anew each time.

If, as this last point suggests, a number  of different senses are stored in memory, one might wonder what part  of the stored  knowledge  should  be identified  as the concept that constitutes  the word’s meaning. One way to resolve the problem  may be simply to say that each word  has  multiple  concepts  associated  with it, with the concepts  in some cases differing only subtly from  one  another. Just  as  a  person’s  concept  of  a sparrow vs. of a finch may be distinct but only slightly different, so may be his or her concepts associated with a polysemous word like ‘run.’

4.    Meaning Across Languages

So far, all the examples discussed have been of English words  and   their  associated   concepts.   Speakers   of English generally have the intuition  that  the distinctions they make between chairs and sofas, bottles and jars, and so on are so obvious  that  the same distinctions must be observed by speakers of other languages: the  words  would  differ  but  the  concepts  associated with them would be the same. However, the situation is  not   that   simple.   Words   that   are   traditionally translated as equivalent by dictionaries are not always fully equivalent  on closer examination. For  instance, Spanish  speakers  apply the word ‘botella’ to a much narrower range of objects than English speakers apply the  word  ‘bottle’  to.  On  the  other   hand,   English speakers  apply  the  word  ‘on’ to  a  more  restricted range of spatial relations than Spanish speakers apply the word ‘en’ to. Apparently, the concepts associated with the words are not identical. Such differences can be  found   across   many   languages   and   for  words labeling many different kinds of entities.

This observation raises at least two important questions, one theoretical and one practical. The theoretical one concerns the relation between language and thought. A controversial proposal  in psychology and  linguistics has  been that  the  language  a person speaks influences the way he or she perceives the world (see  Lucy   1992).  If  the  meaning   of  ‘bottle’  and ‘botella’  are  not  the  same,  or  the  meaning  of  ‘on’ and ‘en’ are not, do people in some way perceive the objects  or  spatial  relations  differently  depending  on their  language?  For  instance,  if five objects  are  all called  ‘bottle’  in  English,  but  require  two  different names in Spanish, does the Spanish speaker see them as more different than  the English speaker does? The available evidence suggests not (Malt et al. 1999). However, the cross-linguistic differences may influence other  aspects  of thought, such as how early a child learns a word or what features he or she first focuses on (e.g., Bowerman  1996).

The practical  issue concerns second language learning. Second language instruction has typically taught vocabulary  as  paired   associate   learning:   students learn   that    ‘chair’   is   ‘silla’   in   Spanish,    ‘bottle’ is ‘botella,’  and  so  on.  If  the  words  are  not  truly equivalent  in their meaning,  then learners’ use of the words may not be like native speakers’. For instance, learners  may use the word  ‘botella’ for the same set of objects that  they would use ‘bottle’ for in English and not realize that their use is inappropriate. It may take  second  language  learners  extensive  experience with the language to fine tune their meanings and their word use to match native speakers, if they ever do.

Bibliography:

1. Bowerman M   1996  Learning   how   to   structure   space   for language:    A   cross-linguistic        In:   Bloom   P, Peterson  M A, Nadel  L,  Garrett M F  (eds.) Language  and Space. MIT Press, Cambridge,  MA, pp. 385–436
2. Clark E V,  Clark  H H  1979  When  nouns  surface  as  Language 55: 767–811
3. Hock H H,  Joseph   B D  1996  Language  History,   Language Change, and Language Relationship: An Introduction to Historical Linguistics. Mouton de Gruyter, Berlin
4. Laird C 1978 Language  and  the dictionary. In: Guralnik D B (ed.) Webster’s New World Dictionary of the American Language, 2nd edn. Collins, Cleveland, OH
5. Lakoff G 1987 Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. The University of Chicago Press, Chicago
6. Lucy J A 1992 Language Diversity and Thought: A Reformulation of the Linguistic Relativity Hypothesis. Cambridge University Press, Cambridge,  UK
7. Malt B C,  Sloman  S A,  Gennari   S,  Shi  M,  Wang  Y  1999 Knowing  versus naming:  Similarity  and  the  linguistic  categorization  of artifacts.  Journal of Memory  and Language 40: 230–62
8. Murphy G L, Medin D L 1985 The role of theories in conceptual coherence. Psychological Review 92: 289–316
9. Nunberg G  1979  The  nonuniqueness of  semantic  solutions: Polysemy. Linguistics and Philosophy 3: 143–84
10. Putnam H  1975  Mind,   Language,  and     Cambridge University  Press, New York
11. Rosch E, Mervis C B 1975 Family resemblances: Studies in the internal structure   of     Cognitive  Psychology   7: 573–605