The Interplay of Synonymy and Polysemy

full
(Murphy 
2003) or 
absolute synonymy
(Cruse 1986). These are words that are identical in meaning 
and can be interchanged in all syntactic contexts. An example would be 
groundhog
and 
woodchuck
. It has often been pointed out that absolute synonyms are extremely rare and 
possibly non-existent. Words will tend to differ in at least some respects or become 
specialized to specific contexts. See Murphy (2003:161-165) for a discussion of the reasons 
languages disprefer full synonyms.
One reason that full synonymy is less common is the fact that many words have 
multiple senses; that is, words are often polysemous. Full synonyms would need to be the 
same across all senses. A second type of synonyms arises when two words have one sense 
where they are the same, but differ in all the other senses that they express. This is called 
sense synonymy
(Murphy 2003) or 
cognitive synonymy
(Cruse 1986). An example is 
father
and 
daddy 
(Cruse 1986:276). Both refer to a biological father. But 
father 
has another sense 
where it refers to a religious figure; 
daddy 
does not refer to a religious figure. For cognitive 
synonyms, there will be semantic similarity and substitutability in one of its senses, but the 
other senses of the words will differ.
A third type of synonymous relationship is called 
near-synonymy
or 
plesionymy
(Murphy 2003, Cruse 1986, Hirst 1995). Near-synonyms have no senses which are exactly 
the same, but the senses are very similar. Murphy (2003) gives the example of 
mob
and 
crowd
. Both words refer to groups of people, but 
mob
usually refers to a crowd that is 
disorderly or potentially breaking the law. Cruse (1986:285) argues that plesionyms produce 
sentences with different truth-conditions. One can deny one word of a plesionym pair, while 
declaring the other: 
That wasn’t a mob, just a crowd.

---

5 | 
These three types of synonyms are not always adopted by linguists. Some conceive of 
synonymy as one end of a continuum of semantic similarity (Miller & Fellbaum 1991:202). 
Under this view, words lying on one end of the continuum would be more synonymous than 
words lying at the other end; potentially capturing the fact that some words are better 
synonyms than others (Murphy 2003:22). Note that the concept of a scale is not incompatible 
with the previous classification of synonym types. Full synonyms can be seen as lying at the 
far end of semantic similarity. Near-synonyms lie further down the scale, with a fuzzy line 
dividing near synonyms from words which are not synonymous (Cruse 1986:268). Still some 
authors adopt the concept of a scale, without using any of the three categories of synonyms.
Adopting a scale of semantic similarity avoids problems that can occur when 
categorizing synonyms into the three types. Placing words into one or another category is not 
always clear-cut. The choice often times depends on several theoretical assumptions that the 
authors adopt. Broadly speaking, these can be divided into two opposing views. In one view, 
there is a strict division between linguistic and extralinguistic elements, meaning is judged 
using truth-conditions and elements such as collocational differences are idiosyncratic and 
irrelevant. Under another view, this strict division does not exist and pragmatic elements exist 
alongside semantic elements to define meaning, truth conditions do not adequately represent 
meaning and collocations signal meaning differences.
Cruse (1986) can be used to exemplify this first position. He relies on truth-conditions 
to decide on synonymy. This was shown with example (1). Using truth-conditions, 
father 
and
 
daddy 
are synonyms. Researchers (Bosque 2004, Goldberg 1995) have noted, though, that 
truth-conditions do not entirely capture meaning as it is intuitively understood. Tuggy 
(1985a) argues that one can define “meaning” in different ways. Cruse (1986) employs what 
Tuggy terms