42 ROBERT D. VAN VALIN, JR.
thematic relations of more agent-like or more patient-like roles. This con
tinuum is flexible, for languages divide it up in different ways. There are,
however, certain restrictions upon the ways that it can be divided up.
Furthermore, the divisions are not random, for there are common divi
sions that we can find across languages. Languages distinguish entities
which are affected in terms of their state or condition from those which are
affected in terms of location. Therefore, they differentiate patients from
things that are clearly at the right end of the scale but are less patient-like,
namely themes. Themes are not affected in the same way and to the same
degree, because they do not get crushed and chopped, but they do get
moved around. Therefore, themes are clearly at the right end of the con
tinuum. On the other hand, there are things on the left side of this con
tinuum which can be causal, but not instigating or controlling. Examples of
this include instruments and forces, which are grouped together under the
broader term, effector. This labels the participant that brought something
about, but there is no implication of its being volitional or the original
instigator. It is simply the effecting participant. Agents are also effectors,
but they also have an additional meaning of control and intent; and they are
the willful initiating participant. To the right, towards the center of the con
tinuum, there is experiencer, which is the locus of an internal event, but is
not willful, volitional, and instigating. Many perceptual events are not voli-
tionally instigated. The difference between verbs like listen to and hear is
that the subject of listen to is an experiencer which is also an agent, while
the subject of hear is simply an experiencer. This is an example of how lan
guages often have pairs of verbs, with one being a volitional and the other
being a non-volitional perceiver. In the middle of the continuum, there are
all of the various types of locatives. In fact, many more can be found in the
world's languages than are represented in Figure 15. Typically things like
recipient, source, goal, and path do not interact with each other, but rather
with theme. The basic claim is that Figure 15 represents a continuum of dis
tinctions, and languages may make more or fewer of them than what is
listed above. This seems to be the semantic range that needs to be expres
sed in all languages. The same distinctions recur across languages; the same
contrasts between experiencer and agent, effector and agent, theme and
locative, and so on, are found in language after language.
With respect to the issue of the number of thematic relations to be
posited in universal grammar, it is necessary to recognize that thematic
relations have traditionally played two distinct roles in syntactic theories: