The Interplay of Synonymy and Polysemy

Table 15
. Results from the questionnaire showing the distribution across the four sentence 
types with goodness-of-fit p-values 
Directionality 
Optionality 
Upward motion 
Downward motion 
Without directional 
With directional 
GOF p-
value 
GOF p-
value 
GOF p-
value 
GOF p-
value 
arrojar 
18 ↓ 
1.24E-04 
59 
↑ 
3.09E-04 
33
0.4193 
45 
0.4492 
echar 
11 ↓ 
4.08E-05 
24 
0.1309 
32
0.6123 
56 
↑ 
0.0003 
lanzar 
143 ↑ 
1.54E-17 
59 
↓ 
8.06E-03 
60 
↓ 
0.0458 
50 
↓ 
0.002 
tirar 
81 ↓ 
5.40E-03 
116 
0.3654 
119 
↑ 
0.0538 
108 
0.9795 
If we look at the upward motion results, the verb 
lanzar
has a very strong preference 
for expressing upward motion, while all the other verbs are used less than expected in 
sentences of this category. The GOF p-value for 
lanzar
is quite small, indicating that the 
results from the questionnaire differ considerably from the expected values. 
Lanzar
in fact 
shows such a high preference for upward motion, that it has statistically fewer occurrences in 
all other categories (downward motion, without directional and with directional). This result 
coincides with the results from the collostructional analysis. Combining 
lanzar
with phrases 
indicating upward motion, including 
al aire
‘to the air’, is preferred to using other verbs.
I argue in the previous section that the collostructional results are collocational 
preferences. Basically, the argument is that it is more correct to say that 
lanzar
attracts the 
phrase 
al aire 
‘into the air’ specifically, and not all phrases that express upward motion in 
general. Superficially, the results from the questionnaire seem to run counter to this 
hypothesis since 
lanzar
does have a generalized attraction to phrases that mean upward 
26
2.20E-16 is the lowest value given by R.

---

118 | 
motion. Recall that three sentences from the upward motion set do not include the phrase 
al 
aire 
‘into the air’. 
To determine which analysis is more accurate, we can compare the totals for 
sentences from the upward motion set that used 
al aire
‘into the air’ to sentences with other 
types of 
DIRECTIONAL 
expressing
upward motion:
encima del closet
‘on top of the closet’, 
al 
techo de la casa
‘to the roof of the house’, and 
al segundo piso
‘to the second floor’
. 
We can 
use the goodness-of-fit (GOF) test to compare the observed data for sentences with 
aire
‘air’ 
and sentences with other phrases. If the choice of phrase is irrelevant, then we should find 
that the observed data should not differ significantly from the expected values. In other 
words, the proportion of 
aire 
sentences to other sentences should not vary across verbs. Table 
16 shows the observed data, the expected values and the GOF p-values.