LINGUIST List 7.1804

Fri Dec 20 1996

Sum: Evaluative verbs

Editor for this issue: T. Daniel Seely <>


  1. Jeff Turley, evaluative verbs

Message 1: evaluative verbs

Date: Thu, 19 Dec 1996 19:02:16 -0700 (MST)
From: Jeff Turley <>
Subject: evaluative verbs

A few weeks ago I posed the following query to _Linguist_:

"I have stumbled across what I believe to be an interesting universal
class of verbs, and would like to know if this category has been
noticed before, or if existing semantic theories can describe it more
precisely than I will here.

The lexical category in question comprises verbs that evaluate certain
events and classify them as being of a certain type, instead of
describing a specific situation type themselves. The classification
does not follow from any property inherent in the event or state. One
of the clearer examples I can give is "help". The verb alone
communicates that the subject does something that is beneficial to
another being and nothing more. How the subject brings about the
beneficial effect must be specified in the rest of the predicate or
must be interpreted from context.

These verbs--let's call them "evaluatives"--create generalizations. 
For instance, saying that a friend helped me by covering my shift at
work effectively places the action (i.e., covering my work-shift) in a
class of beneficial actions. Although subordinate-level verbs (e.g.,
"murder" vs. the basic-level "kill") are also evaluative ("murder" is
a kind of killing), I don't think evaluatives are simply equivalent to
subordinate level verbs, since the latter always describe the same
general situation type as their corresponding basic-level counterpart,
only with greater precision. But an evaluative verb may depict more
than one situation type, even totally different situation types. If I
am really hot, I could say "Would you help me by pouring this bucket
of cold water on my head?." But if I was dressed up for a party and
you poured water on my head, I wouldn't say that you had helped me,
unless of course I was going to a party where everyone was supposed to
wear wet party clothes. What is considered helpful is relative to the
context. Other verbs that seem to be evaluative include use, harm,
pamper, care for, be careful, prepare, defend, neglect, ruin,
discredit and use."

I received some extremely helfpful comments, many more than I had
expected, given the esoteric nature of my query. Thanks go to Suzanne
Fleischman, Gregory Ward, Rebecca Wheeler, Marco Ruehl, David
Robertson, Doris Marszk, and one other person whose message I
accidentally erased (thanks to you--you know who you are).

As I had hoped, and perhaps feared, I am not the first to notice these
verbs. Back in 1951, Gilbert Ryle coined the term "polymorphous",
which is a slightly more general category than my "evaluatives." 
Summarizing Ryle, Alan White, in _Attention_ (Blackwell, 1964, pp.
5-6), describes polymorphous verbs as follows: "An activity-concept X
is `polymorphous,' that is, one which takes many forms, if there are
many different activities the engagement in one or other of which can
in certain circumstances count as X-ing, and yet none of which
activities in other circumnstances necesarily counts as X-ing. It is
the second of these two conditions which distinguishes a polymorphous
concetp, like _working_, from a generic concept, like _perceiving_." 
A paradigm example of a polymorphous verb is _repeat_. (Thanks to
the unkown listmember who pointed me in Ryle's direction.)

Several other respondents emphasized that there was a kind of modality
involved with evaluatives. This modality could be characterized along
the lines of factivity, for example. Marco Ruehl made an interesting
connection to work he is pursuing in argumentation theory/analysis. 
Springboarding from the work of Christian Plaintin (Essais sur
l'argumentation, Paris, Kime, 1990), Marco distinguishes between
"auto-argumentivity" and "hetero-argumentivity". The former refers to
laudative, vituperative, or otherwise classifying terms, over against
"neutral" ones, that speakers use during an argument. Such terms by
their very meaning beg the question of the argument, by characterizing
the topic under discussion. 

Gregory Ward tipped me off to his fascinating study (in collaboration
with Brian Bowdle) on "generic demonstratives" (e.g., those IBM
ThinkPads sure are expensive!). They found that the relevant
predicate must be evaluative (*those IBM ThinkPads are black), where
the subject referent is generic. Thus, the principle of evaluation
has grammatical consequences.

Finally, Doris Marszk graciously informed me that she has already
discovered these verbs; in fact, they constitute her primary area of
research: "Granularitaet als lexikalische Kategorie sui generis," in
Uwe Junghanns, ed., Linguistische Beitraege zur Slawistik aus
Deutshland und Oesterreich, II, JungslavistInnen-Treffen, Leipzig,
1995, Vienna: Wiener Slawistischer Almanach Sonderband 37, 1995, pp.
205-217. and "Russische Verben und Granularitaet," Specimina
philologiae slavicae, Supplementband 47, Munich, 1996.

Doris approaches the question from a cognitive perspective. She dubs
these verbs "Auslegungsverben" (interpretative verbs), which I like
better than "evaluatives". I'm just going to forward her summary: 

> I found these verbs while working on my dissertation project about
> granularity as a lexical category of verbs. Interpretation-verbs
> which I found are: to help, to save, to harm s.o., to fail, to
> support and some others. But I investigated these verbs in Russian
> and in German. It may be that not all the verbs I found are found
> in other languages, e.g. in English, as interpretation verbs. For
> instance, the german verb "retten" or the russian verb "spasti"
> (both "save") can be used in nearly infinitely many situations. I
> don't know whether this is the case with "save" in English, too.
> Interpretation-verbs are in my system a special group of "coarse
> grained verbs". I assume that verbs have a certain "granularity",
> i.e. they are, so to speak, more or less concrete or abstract. The
> more concrete ones I call "fine grained", the more abstract ones I
> call "coarse grained", but there are several gradations. Fine
> grained verbs are e.g. "to pour", "to roll up", "to press", "to
> push" etc., coarse grained verbs are e.g. "to emigrate", "to
> travel", "to study", "to cure" etc.

In parting, some further reflections of my own. It seems to me that
these verbs involve differing degrees of interpretation or construal. 
Those that require a high degree of construal are those that make
emotionally charged or otherwise controversial judgments, such as the
evaluations of performance (fail, screw up, do a great job). Many of
this last kind involve verbs of speaking--bore, fascinate, wander,
drone on, brag, gab, classify speech events. Other verbs equate a
speech event with a kind of speech act, such as `lie', `protest' or
`slander': there is no way to predict the propositional content of
what might be construed as a protest, for instance. On the other
hand, a verb like `chatter' requires less construal. 

In any event, many thanks to all of you. I am encouraged to pursue
this topic because of the response I received. 



Jeffrey S. Turley

Department of Spanish and Portuguese
Brigham Young University

Wisdom comes from good judgment.
Good judgment comes from experience.
Experience comes from bad judgment.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue