Rit (Vísindafélag Íslendinga) - 01.06.1970, Side 256

254
Homograph COMPONENT Frequency Homograph COMPONENT Frequency
i PP 29,461* han PN 6,446*
en AL 18,184* som KN 6,079*
ár VB 15,506* de AL 5,695*
av PP 15,242* men KN 5,677*
det PN 13,991* sig PN 5,559*
pá PP 13,932* den PN 5,378*
som PN 13,522* om PP 5,'99*
att IE 12,534* de PN 4,997*
för PP 11,807* kan VB 4,479*
med PP 11,700* det AL 4,203*
att KN 11,232* frán PP 4,042*
till PP 9,865* sá AB 3,979*
den AL 9,290* var VB PRT 3,839*
har VB 9,176* om KN 2,283*
ett AL 7,702* en NL 2,101*
man PN 7>I03*
Table 4. Frequency of the Commonest Components
of the Most Common Homographs
IE for the sign of the infinitive, KN for conjunction, AB for ad-
verb, PRT for preterite, and NL for numeral. Asterisks indi-
cate estimated frequency (based on sampling). The list of ab-
breviations just given indicates the dominance of structure
words. An interesting fact not shown in the table is that the
noun man ‘man’, the pronoun var ‘each’, the verb för ‘convey
(pres. and imper.)’, the verb sá ‘sow’, and the noun men ‘disad-
vantage’ are among the low-frequency components of the ho-
mographs listed. Items like these would get lost, if the most
frequent words were simply to be cut off from the separation
of homographs.
Lemmas. Count no. 4 of Hassler-Göransson concerns the
vocabulary of newspapers in 1932/33. This makes possible,
in certain respects, a comparison with the present investiga-
tion. Let us consider the subsystem of ecclesiastical words.
First we notice that words of this kind are clearly over-
represented in the conservative press; see Table 5 and cf.