Synonymy and length - Revision history

Ahans at 12:28, 11 July 2006

2006-07-11T12:28:25Z

Ahans at 10:30, 3 July 2006

2006-07-03T10:30:54Z

Ahans at 07:33, 31 May 2006

2006-05-31T07:33:04Z

New page

'''1. Problem and history'''

The number of synonyms of a word is indirectly associated with its length (if length is variable in the given language). This is the consequence of the fact that both length and synonymy depend on polysemy, the former in negative the latter in positive sense. The specification of meaning, i.e. reduction of polysemy, leads to a prolongation of the word (by affixation, compounding, reduplication etc.) and to the reduction of possible synonyms. Thus we have the scheme:

The only hypothesis set up and tested is that of Wimmer and Altmann (2001a) concerning Italian synonyms. Corroborating results brought Uhlířová (2001a) for Czech and Rottmann (2001a) for Russian, Bulgarian, Polish and Ukrainian.

More complex dependencies have not been tested as yet.

'''2. Hypothesis'''

''The number of synonyms of a word is a function of its length (or vice versa)'' .

'''3. Derivation'''

Using the unified theory (see § 4.1, Example 1) we assume that the relative rate of change of synonymy is proportional to the relative rate of change of length while polysemy plays an intermediary role expressed by a constant c, i.e. (y = synonymy, x = word length)

(1) <math> \frac{dy}{y}= \left( c+\frac{b}{x} \right)</math>

resulting in

(2)<math> y= Ax^b e^{cx}\quad</math> .

This concerns, of course, the ''mean number of synonym''s for all words of a given length.

'''Example'''. Length and synonymy in Italian

Wimmer and Altmann (2001a) took a sample from the Italian dictionary by P. Stopelli, Sinonimi e Contrari con generici, Spezifici, Analoghi, Inversi. (Milano, Garzanti 1998) and found the results presented in Table 1 and Figure 1. As can be seen, the curve is here concave.

Table 1
Dependence of synonymy on length in Italian (Wimmer, Altmann 2001)

Word lengthx Number of words with length xfx Number of synonyms of all fx words in class xnx Mean number of synonymsyx = nx/fx Theoretical values (2)
12345678 80 95 89200114146 54 4 282 705 6341364 625 915 207 10 3.537.427.126.825.486.143.832.50 3.916.667.487.076.094.953.852.91
A = 6.2752, b = 1.4513, c = -0.4731, D = 0.88

<div align="center">[[image:Grafik1_SaL.jpg]]</div>
<div align="center">Fig. 1. Dependence of mean synonymy on word length</div>

Further testing is necessary. It must be noted that in the observed data only in one case (Bulgarian, cf. Rottmann 2001a) the value of x = 1 was greater than the other ones.

'''4. Authors: U. Strauss, G. Altmann'''

'''5. References'''

'''Rottmann, O'''. (2001a). On the “second law of synonymy”: observations in Russian, Bulagrian, Polish and Ukrainian. In: Ondrejovič, S., Považaj, M. (eds.), ''Lexicographica ´99'': 251-257. Bratislava: Veda

'''Uhlířová, L.''' (2001a). Kolik je v češtine synonym? (K dynamické stabilitě v systému lexikálních synonym). In: Ondrejovič, S., Považaj, M. (eds.), ''Lexicographica ´99: 237-250''. Bratislava: Veda.

'''Wimmer, G., Altmann, G'''. (2001a). Two hypotheses on synonymy. In: Ondrejovič, S., Považaj, M. (eds.), ''Lexicographica ´99: 218-225''. Bratislava: Veda.

← Older revision		Revision as of 12:28, 11 July 2006
Line 29:		Line 29:
	Wimmer and Altmann (2001a) took a sample from the Italian dictionary by P. Stopelli, Sinonimi e Contrari con generici, Spezifici, Analoghi, Inversi. (Milano, Garzanti 1998) and found the results presented in Table 1 and Figure 1. As can be seen, the curve is here concave.		Wimmer and Altmann (2001a) took a sample from the Italian dictionary by P. Stopelli, Sinonimi e Contrari con generici, Spezifici, Analoghi, Inversi. (Milano, Garzanti 1998) and found the results presented in Table 1 and Figure 1. As can be seen, the curve is here concave.

−		+	<div align="center">[[image:Tabelle1_SaL.jpg]]</div>

@@ Line 2: / Line 2: @@
 The number of synonyms of a word is indirectly associated with its length (if length is variable in the given language). This is the consequence of the fact that both length and synonymy depend on polysemy, the former in negative the latter in positive sense. The specification of meaning, i.e. reduction of polysemy, leads to a prolongation of the word (by affixation, compounding, reduplication etc.) and to the reduction of possible synonyms. Thus we have the scheme:
 The only hypothesis set up and tested is that of Wimmer and Altmann (2001a) concerning Italian synonyms. Corroborating results brought Uhlířová (2001a) for Czech and Rottmann (2001a) for Russian, Bulgarian, Polish and Ukrainian.
@@ Line 29: / Line 31: @@
 <div align="center">[[image:Grafik1_SaL.jpg]]</div>