The 1992 Presidential Debates:
A Cognigraphic View
Richard F. Carter Keith R. Stamm
Professor Professor
School of Communications DS-40 School of Communications DS-40
University of Washington University of Washington
Seattle, WA 98195 Seattle, WA 98195
(206) 543-9676 (206) 545-0127
Paper submitted to Communication Theory & Methodology
Division, AEJMC, 1993.
ABSTRACT ABSTRACT ABSTRACT
The 1992 Presidential Debates:
A Cognigraphic View
The 1992 presidential campaign, especially its presidential
debates, produced cognitive, affective, and emotive effects which
the Cognigraphics research method was able to distinguish. George
Bush's campaign decline was dramatically etched in respondent
cognitions. William Clinton survived early emotional and
continuing affective disadvantage, even after the debates, to win
the election on what seemed clearly cognitive grounds -- the
country's need for a change by someone seen as capable of
changing things. Ross Perot was strongly criticized for dropping
out of the race, but his debates performance turned around his
affective picture, even though respondents seemed unconvinced
that he was capable of making the needed changes.
The 1992 Presidential Debates:
A Cognigraphic View
Who won the 1992 presidential debates? George Bush? William
Clinton? Ross Perot? In what sense could these debates be said to have
been won? Such questions are not easily answered. The three 1992
presidential debates presented a thicket of conditions with many
potential effects. Further, they were embedded in an election campaign
of even greater length and breadth. Assessing their effects is no easy
matter. Four avenues suggest themselves:
* As usual, we could rely on analyses from multiple observers
taking various perspectives. These might provide a consensus view of
at least the most obvious of the debates' significance.
* Also as usual, we could wait on future events and on
reinterpretations consequent to those events for a fuller assessment
of the debates' significance. This too would be biased toward the most
obvious (i.e., persisting) indications of significance.
* We could arrange to get an inside-the-debates look at moment-
to-moment reactions to them, experimentally exposing nonviewers to
them, such as by using the "signaled stopping" research technique
(Carter et al, 1974; Hawkins et al, 1991). This approach has the
virtue of getting closer to the "when" of effects. Less obvious and
less lasting effects are also not so likely to be missed.
* We could arrange, via "before" and "after" surveys, to get a
reading from viewer and nonviewer responses to the debates. This
concern for the "what" of effects is limited by what we arrange to
look for. What kinds of effects do we expect such debates to have? Can
- -
our methods help us find such effects if they are there to be found?
The Cognitive Challenge
Attitudinal and participation (state) variables are often able to
indicate whether or not some media effect has occurred, but they have
not been very helpful in telling us much about the nature of media
effectiveness. This has led reviewers to call for greater attention to
the media's cognitive effects (e.g., Chaffee and Hochheimer, 1985 ;
Rogers and Dearing, 1988).
Further, in electoral contexts, vote prediction via measures of
affect has often held sway over understanding of cognitive effects. In
typical attitudinal research, affect (re candidate preference) serves
as an indicator of disposition or predisposition (Green, 1954).
Cognition, narrowly interpreted as knowledge, only contributes to the
attitudinal complex (Katz, 1960). We have Behaviorism to thank for
this view. Having limited behavior to the body's movement toward or
away from other objects, the concept of attitude was given the job of
representing whatever led to that directed movement. Hence we have
attitude's ambivalent and tautological specification as either
disposition or predisposition.
The key to a better understanding of media effectiveness,
potential as well as accomplished, seems to lie in our doing a better
job of getting into the minds of media users. How do they mind? The
Behaviorists' impoverished conceptualization needs dispelling.1
The question of understanding media effectiveness is crucial to
the functional role -- and the future -- of mass media as servants of
democratic society. If we are able to demonstrate only an occasional
fact of this or that media effect but unable to comprehend the nature
of media effectiveness, how are we to design the media we need for our
polity's success and survival?2
This study reports how the three presidential candidates appear
to have come across in the debates. We used a new methodology,
Cognigraphics, to register and analyze what was uppermost in
respondents' minds. Additional data from earlier Cognigraphic surveys
help interpret these results in the electoral context.
***
Methods
Before and after the 1992 debates, we surveyed students enrolled
in a University of Washington undergraduate "Communication Process"
course. Most of these students are not communication majors. The
course meets a UW distribution requirement. We had surveyed cohorts of
these students previously in this election year, early in the primary
season (when Clinton was one of five major Democratic nominees) and
later in the primary season (when Perot had become a contender). We
also have cohort data on Bush from surveys conducted in 1991 during
the Gulf War period.
The Cognigraphics method weds the Freudian word association
tactic to a simple form of ideational mechanics.3 The respondent first
gives a word associate for each presented topic (here, the three
candidates), then chooses the PIX option which best represents the
relation of the topic to the word associate. The PIX options are
presented to respondents as drawings:4
#1: Topic outside the word associate
#2: Word associate outside the topic
#3: Topic before/makes difference in word associate
#4: Word associate before/makes difference in topic
#5: Topic same as word associate
#6. Topic not same as word associate
Using the topic of George Bush and the word associate of "president,"
examples of the six PIX are:
#1: Bush has president as an attribute
#2: Bush is an instance of a president
#3: Bush makes a difference in the presidency
#4: The presidency makes a difference in Bush
#5: Bush and president are the same thing
#6: Bush is no president
A topic-word associate relationship can be reasonably interpreted
several ways. So, one key to what people think is how they think. And
it is clear that what people might come to think together will depend
on our ability to make apparent how they are thinking about their
common focus of attention.
The significance of thinking about a topic one way or another is
a matter of some importance, as our previous work with Cognigraphics
has shown. For example, all six of the Bush examples above come from
respondent usage. An uninterpreted (topic-word association)
relationship does not tell us enough about how people think. The use
of PIX #3 often suggests greater topic consequentiality (Carter,
Stamm, and Heintz-Knowles, 1993); the use of PIX #5 often suggests
heightened respondent emotion, because it fails to make a distinction
when common usage of the word associate implies such applicability
(Carter and Stamm, 1993), In our presentation of the results we shall
discuss further apparent implications.
Respondents' word associates also contribute to our analysis. The
most frequent word associates tell us something. We can also code them
for evidence of affect. In light of the thin slice we get from the top
of cognitive activity, we consider this surface affect. This is not
just to distinguish it from the more familiar attitudinal affect as
measured, say, by Likert scales, but to remind us that expressing such
content can elicit a response of, "Those are fighting words." Because
this is freely expressed affect, we also have the advantage of not
having solicited an expression of affect. Thus we can usefully view
the ratio of affective to nonaffective word associates.
Examples of word associate and PIX usage were shown to students
in class via overhead projection. Additional questions about the
debates and election were then asked. These student cohorts do not
constitute a random sample of the American voter population, but their
responses do help us discover something about the nature of
cognizing.5
***
Results and Discussion
George Bush
Polls had been showing a decline in popularity for Bush as public
attention turned from the Gulf War to the state of the domestic
economy in an election year. But a response to a popularity poll, like
the electoral vote, can be based on a variety of cognitions -- or
perhaps none at all. What was going on cognitively? Would the debates
change ideas about Bush?
We had begun auditing Bush Cognigraphically during the Gulf War,
at the time of the United Nations entry, after we had obtained very
unusual results for Saddam Hussein as a topic (Carter and Stamm,
1993). Respondents assigned Hussein quite a variety of word
associates, many of them negative in affect (e.g., "mad," "dictator").
Whatever the word associate, however, they distinctively chose PIX #5.
They equated him with these conditions; they did not view them as
attributes (PIX #1). Very interestingly, respondents in the next
cohort continued this practice and they also most often chose PIX #5
for Bush. Here, however, there was relatively little variety of word
associate. "President" dominated. And affect was largely absent. We
concluded that use of the #5 PIX indicated emotion due to the Gulf
War, emotion which seemed to get in the way of more discriminating
thoughts about these two war leaders.6
Table 1 shows our studies' history of "president" as a word
associate for Bush. From the height of his post-war popularity to the
beginning of the debates and on to their conclusion we see a sharp
decline. Observer consensus before and after the election was that
change would, and did, win. His failure to make domestic changes
appears to have eroded his presidentiality.
----------------------------------------------------------------------
Table 1. "President" as word associate for Bush: 1991-1992.
Survey_date Proportion_of_word_associates
1/23/91 (N=111) 43.5%
3/14/91 (N=169) 64.3
2/12/92 (N=101) 37.6
10/7/92 (N=139) 30.1
10/22/92 (N=161) 22.8
Chi square = 53.4; 4 df; p < .001
----------------------------------------------------------------------
Table 2 shows that word associates for Bush were not very affect-
laden, either before or after the debates. Clinton and Perot, we shall
see, drew more affect. Affect for Bush was down after the debates.
Though not statistically significant, we can hardly miss the
intimation that fewer people seemed to care one way or the other about
him. What there was of affect was still about three to one negative
over positive, which is clearly a significant difference. On positive
affect, Bush benefitted very little from his debates performance. As
we shall see later, Clinton and Perot, especially Perot, benefitted
more from their debates performances.
----------------------------------------------------------------------
Table 2. Affective word associate usage for Bush, before and after
the debates.
Affective usage (in % of all word associates):
Time Negative Positive Total_affect
Before (N=139) 26.6% 6.5 33.1
After (N=161) 21.8 8.1 29.9
Chi square = 1.11; 2 df; p = n.s.
----------------------------------------------------------------------
Table 3 shows that Bush PIX usage changed in two respects from
before to after the debates. First, there was increasing use of #4
PIX. Differences were seen as being made in Bush instead of his making
differences. In previous studies, the topic of environment elicited
this kind of an idea. Things are thought to happen to it rather than
it making something happen -- an important policy consideration.
Second, #2 PIX usage decreased. This usage often suggests less
topic significance than, say, #1 PIX usage. So the debates might have
contributed to rescuing him from consignment to a category (that of
ex-presidents, judging from the word associates).
We have also included PIX data from the Gulf War period to
provide added context. Emotion (#5 PIX) had subsided from that level
by the time of the debates. It did not rise much afterwards.
----------------------------------------------------------------------
Table 3. PIX usage for Bush before and after the debates, and during
the Gulf War.
PIX: (in %)
_#1_ _#2_ _#3_ _#4_ _#5_ _#6_ Total
1/23/91 (N=111) 29.7 10.8 16.2 3.6 36.9 2.7 99.9
Before (N=135) 20.7 20.7 21.5 8.1 26.7 2.2 99.9
After (N=151) 19.9 10.6 22.5 14.6 29.8 2.6 100.0
Chi square = 21.6; 10 df; p < .01
----------------------------------------------------------------------
Table 3 affords us an opportunity to introduce a measure that we
calculate from PIX usage. We call it the "Power Ratio." We add the #1
and #3 PIX usages and divide them by the #2 and #4 usages. We consider
being included (#2) and being made a difference in (#4) evidence of
less consequentiality than including (#1) and making a difference in
(#3). On this measure, Bush was impressive when we entered the Gulf
War: 3.19 (45.9% divided by 14.4%). But before the debates he was down
to 1.47. And after the debates he was not much better -- 1.68,
trailing both Clinton and Perot on this measure.
***
William Clinton
Clinton had one advantage of sorts as the primaries began. About
half of our respondents could not at that time even furnish a word
associate for the other four Democratic candidates. Most could for
Clinton. However, the most frequent of these were "scandal," "affair,"
"promiscuous," and so on (53.3% in all). One reason these students
might have had so much of this on their minds was a discussion on
recent tabloid coverage of Clinton in a previous class session.
When we surveyed another cohort in May, 96.6% of the respondents
could furnish a word associate for Clinton. The sexual word associates
had all but disappeared. "Democrat" (9.8%) and "slick" (7.6%) were the
most frequent word associates. They still were just before the debates
(10.4%; 7.6%, respectively), but "change" (4.9%) had joined them. The
same three were most frequent after the debates, but their rankings
were different ("slick," 11.9%; "change," 8.2%; "Democrat," 6.9%).
Table 4 shows that Clinton drew much more affect than Bush. We
also see that the debates seemed to trigger a spurt in his affective
word associates, and that this affect still tended to be negative. He
did get something of a boost in positive affect from the debates.
Clinton had elicited nearly as much affect in May as he did after the
debates. Then, however, it was predominantly negative (46.7% to 4.3%)
----------------------------------------------------------------------
Table 4. Affective word associate usage for Clinton, before and after
the debates.
Affective usage (in % of all word associates):
Time Negative Positive Total_affect
Before (N=139) 25.2% 9.4 34.6
After (N=161) 32.9 19.3 52.2
Chi square = 10.7; 2 df; p < .01
----------------------------------------------------------------------
When the primaries started, as we can see in Table 5, Clinton was
not just making differences (PIX #3), differences were being made in
him (PIX #4). The latter were the scandal and tabloid at work.
Interestingly, the #4 PIX had all but disappeared by the time the
primaries were over. But there is considerable indication of emotion
(PIX #5),7 only some of which is related to the very negative affect
of that same period.
The post-primary emotion appears to have subsided by the time the
debates began, to be somewhat stimulated by the debates, but not so
much as affect was, even though the two are significantly correlated
(phi = .18 before; phi =.23 after). There is no evidence that this
post-debate increase in emotion was due to partisanship, because even
though "Democrat" as a word associate increased slightly, these
relationships were not interpreted with #5 PIX choices.
While Bush's Power Ratio increased only slightly after the
debates, Clinton's surged upwards, from 1.83 to 2.92. Previously, in
May, this measure had also been high (2.84), as he apparently
recovered from the Gennifer Flowers exposure.
----------------------------------------------------------------------
Table 5. PIX usage for Clinton before and after the primaries, and
before and after the debates.
.
PIX: (in %)
_#1_ _#2_ _#3_ _#4_ _#5_ _#6_ Total
2/12/92 (N=80) 16.2 7.5 26.2 26.2 17.6 6.3 100.0
5/14/92 (N=81) 22.2 12.3 19.8 2.5 39.5 3.6 99.9
Before (N=136) 25.7 19.9 24.3 7.4 18.4 4.4 100.1
After (N=146) 26.0 11.0 28.1 7.5 24.7 2.7 100.0
Chi square = 51.8; 15 df; p < .001
----------------------------------------------------------------------
***
Ross Perot
Perot's "out and back in" candidacy helped produce some of our
strongest findings. His dropping out antagonized many, and he came
to the debates handicapped in this respect. But the debates,
particularly the last two, served him well. If change were to win the
1996 presidential election, as it seems to have won this one, it could
be his turn.
The most frequent word associate for Perot was some variant of
"rich." This ranged from 29.3% in May, when we first included him, to
20.5% just before and 18.8% just after the debates. Before the
debates, the next most frequent word associate was "quitter" (8.6%).
Afterwards, it was "businessman" (15.8%) -- and "quitter" was nearly
gone (1.2%).
As Table 6 shows, affective response to Perot was minimal in May,
but relatively high before the debates. Like Clinton in May, his
predebate affect was almost all negative (as "quitter" suggests). Some
of the total affect dissipated with the debates. Most strikingly, his
post-debates affective response took a decided turn toward the
positive.
----------------------------------------------------------------------
Table 6. Affective word associate usage for Perot, before and after
the debates and in May.
Affective usage (in % of all word associates):
Time Negative Positive Total_affect
5/14/92 (N=75) 9.3 9.3 18.6
Before (N=139) 45.3 4.3 49.6
After (N=161) 17.4 18.0 35.4
Chi square = 31.26; 2 df; p < .001
----------------------------------------------------------------------
We found that the number of debates watched did not correlate
significantly with either affect or PIX usage. (There is, however, an
obvious tendency for those who watched the most often to favor Perot
and for those who watched the least to favor Clinton.) Table 7 shows,
in the correlation between watching particular debates and affect,
that Perot clearly benefitted from the last two debates.
----------------------------------------------------------------------
Table 7. Affective word associates for Perot by viewership of third
and fourth presidential debates (in frequencies).
Third debate
Didn't_view Viewed
Positive: 7 22
Negative 14 14 Chi square = 4.1
p < .05
Fourth debate
Didn't_view Viewed
Positive: 11 18
Negative 20 8 Chi square = 6.5
p < .03
----------------------------------------------------------------------
Table 8 shows us that considerable emotion (PIX #5) remained in
October, before the debates, from Perot's summer withdrawal from the
presidential campaign. Just over half of this accompanied negative
affect in a variety of word associates, of which "quitter" was but one
of many -- and not the strongest. Epithets are not uncommon word
associates when negative affect is wedded to emotion.
The emotion appears to have dissipated with the debates. Perot's
performance would seem to have been instrumental. Respondents turned
to PIX #1 and #2 usages instead.
Perot, we see, never attained the PIX #3 usage that Bush and
Clinton did. Thus, while his Power Ratio is relatively high after the
primaries (2.83), this is due primarily to PIX #1 usages. By the time
of the debates he had lost ground (2.42), and the debates evidently
did nothing to improve that situation (2.00). In this regard, Clinton
seems to have been seen as the more appropriate instrument for
effecting needed change.
----------------------------------------------------------------------
Table 8. PIX usage for Perot after the primaries, and before and after
the debates. .
PIX: (in %)
_#1_ _#2_ _#3_ _#4_ _#5_ _#6_ Total
5/14/92 (N=67) 40.3 16.4 10.4 1.5 28.4 3.0 100.0
Before (N=134) 29.9 11.2 17.2 8.2 31.3 2.2 100.0
After (N=149) 37.6 16.8 16.1 10.1 18.1 1.3 100.0
Chi square = 15.42; 10 df; p < .10
----------------------------------------------------------------------
***
In Conclusion
Cognigraphics' word associates have shown us what was cognitively
linked to the three 1992 presidential candidates, and of this how much
was affective. Its PIX options have shown us more of that cognition
and also allowed us to see where consequentiality and emotion played a
part.
George Bush seems to have become less and less presidential after
the Gulf War. He was also seen as less and less consequential. The
debates seem to have done him and his candidacy very little good.
Negative affect plagued William Clinton all the way through the
primaries and even after the debates. But he was seen as more and more
consequential, a reasonable tool for bringing about change. We might
well conclude that cognition, not affect, elected Clinton.
The debates might not have done Ross Perot's candidacy any good,
as is indicated by the degree and kind of consequentiality shown in
PIX usage. But his improvement in affect from before to after the
debates, which can be traced to his performance in the last two
debates, augers well for his credibility as a future candidate.
Finally, taken in conjunction with previous studies, these
results suggest two conclusions. First, Cognigraphics can be a useful
research tool for sorting out and exploring kinds of ideas and several
conditions (affect and emotion) closely related to their expression.
These conditions can make distinctive contributions and interact in
various ways. They are not usefully compounded and confounded as
attitudinal disposition. Second, in these debates much happened
cognitively that needs to be understood if we are to adequately
assess, and then improve, media performance in a democratic society.
##
Endnotes
1. The Behaviorist antipathy toward the cognitive sector of behavior
still finds many adherents among mass communication researchers.
Perhaps not all that conscious of their prejudice-by-adoption, they
limit their conception of behavior to adient and abient movements of
the body relative to other bodies. They must then try to understand
cognition as somehow not "behavior" (only as a component of the
hypothetical summary state variable, "attitude") when cognizing is, in
fact, a crucial and structurally specifiable part of the observing
sector in human behavior, comprising not only ideational thinking but
such important behavioral features as cognitive nonsingularities
(which control starting and stopping).
2. It is one thing to show that the mass media have something to do
with what their audiences pay attention to. See "agenda setting" (Shaw
and McCombs, 1972; and many others). It is something else to know what
additional services they (or others) need to provide the public
(Lippmann, 1922; Lasswell, 1948; Shaw and Martin, 1993; Carter, Stamm,
and Heintz-Knowles, 1993). Ever since the pioneering 1960 presidential
debates, we have been asking if the debates are worth doing and how we
might do them better.
3. Where Freud used word association to dig into an emotion-laden
past, we use it to study the "how" and "what" of current thinking, in
anticipation of improving the "how" of future collective thinking. See
Carter, Stamm, and Heintz-Knowles, op. cit. Ideational mechanics
comprises the basic kinds of cognitive elements and relations people
use to describe what was and what is and to imagine what might be.
(See: Carter, 1991.) Because to share ideas we must be able to
understand each other's ideas, such tools as ideational mechanics for
sharing ideas are readily adaptable to the measurement of ideas held
-- in this case by our respondents about the presidential candidates.
(See: Carter, 1992.)
4. These six statements describe what the PIX look like. Each PIX
covers a number of possible verbal equivalents. The six PIX are used
instead of the 38 (so far) alternative verbalizations. (See Appendix
A.) We ask respondents a third question, "Why did you select that
PIX?" when we want closer specification of the idea represented. This
not only clarifies the PIX usage; it sometimes reveals that the topic-
word associate relationship is part of a larger idea, and thereby
occasionally enables us to withdraw our incredulity that such an idea
could be held. Overall, such questioning casts grave doubt on any
assumption that thinking is innately linked to language usage. The
ideas we find are visual as well as verbal. Expectations of PIX usage
for a given word associate based on such syntactic familiars as
adjective-noun are often proved misleading by our data. For example, a
familiar "adjective" might be assigned PIX #3 instead of PIX #1 if the
topic (say, Clinton) had demonstrated in the debates a capability (to
make this difference maker).
5. Perhaps the best analogy for this technique is the physicist's
bubble chamber, which made it possible to see a variety of collision
consequences. Their interpretation, along with the subsequent
development of much more sophisticated observational contexts,
depended upon a parallel development in theory. The progeny of
Cognigraphics should be psychlotrons.
6. Parallel surveys of Fidel Castro and Mikhael Gorbachev as topics
revealed much different PIX profiles. For them and their word
associates, #2 PIX were chosen significantly more often. (See: Carter
and Stamm, op. cit.)
7. Freud might have been fascinated that an emotional state was to be
found after what might have been the precipitating event's cognitive
content had disappeared from the top of the mind.
***
References
Carter, R. F. (1991). Ideational mechanics. Unpublished manuscript.
Carter, R. F. (1992). Cognigraphics: taking the measure of ideas.
Paper presented to American Association for Public Opinion
Research.
Carter, R. F., Ruggels, W. L., Jackson, K. M., and Heffner, M. B.
(1973). Application of signaled stopping technique to communication
research. In: P. Clarke, ed. New Models for Communication Research,
pp. 15-44.
Carter, R. F. and Stamm, K. R. (1993). How we thought about the Gulf
War. In: B. L. Greenberg and W. Gantz, eds., Desert Storm and the
Mass Media. Cresskill, NJ: Hampton Press.
Carter, R. F., Stamm, K. R., and Heintz-Knowles, K. (1993). Agenda-
setting and consequentiality. Journalism Quarterly, 69/4:868-77
Chaffee, S. H. and Hochheimer, J. L. (1985). The beginnings of
political communication research in the United States: origins
of the "limited effects model." In M. Gurevitch and M. R. Levy,
eds., Mass Communication Review Yearbook, Vol. 5. Beverly
Hills: Sage.
Freud, S. (1953). A General Introduction to Psychoanalysis. New
York: Permabooks.
Green, B. F. (1954). Attitude measurement. In G. Lindzey, ed.,
Handbook of Social Psychology, Vol. 1, pp. 335-69. Reading, MA:
Addison-Wesley.
Hawkins, R. P., Pingree, S., Fitzpatrick, M. A., Thompson, M., and
Baumann, I. (1991). Implications of concurrent measures of viewer
behavior. Human Communication Research, 17:485-504.
Katz, D. (1960). The functional approach to the study of attitudes.
Public Opinion Quarterly, 24:163-205.
Lasswell, H. D. (1948). The structure and function of communication
in society. In L. Bryson, ed. The Communication of Ideas, New York:
Harper & Brothers.
Lippmann, W. (1922). Public Opinion. N.Y.: Harcourt, Brace & Co.
McCombs, M. E. and Shaw, D. L. (1972). The agenda-setting
function of mass media. Public Opinion Quarterly, 36:176-87.
Rogers, E. M. and Dearing, J. W. (1988). Agenda-Setting research:
where has it been, where is it going? In J. Anderson, ed.,
Communication Yearbook 11, pp. 555-594. Newbury Park, CA: Sage.
Shaw, D. L. and Martin, S. E. (1993). The function of mass media
agenda setting. Journalism Quarterly, 69/4:902-20.
Appendix A
PIX and Verbalizations
Some common verbalizations can be seen as instances of the
various PIX....
1. W.A. inside of T 2. T inside of W.A.
1.1. W.A. is quality of T 2.1. T is quality of W.A.
1.2. W.A. is property of T 2.2. T is property of W.A.
1.3. W.A. is component of T 2.3. T is component of W.A.
1.4. W.A. is member of T 2.4. T is member of W.A.
1.5. W.A. is example of T 2.5. T is example of W.A.
1.6 W.A. is development in T 2.6. T is development in W.A.
1.7. W.A. is to be found in T 2.7. T is to be found in W.A.
1.8. W.A. is essence of T 2.8. T is essence of W.A.
3. T comes before W.A. 4. W.A. comes before T
3.1. T makes a difference in W.A. 4.1. W.A. makes a difference in T
3.2. T produces W.A. 4.2. W.A. produces T
3.3. T helps produce W.A. 4.3. W.A. helps produce T
3.4. T is necessary for W.A. 4.4. W.A. is necessary for T
3.5. T is catalyst for W.A. 4.5. W.A. is catalyst for T
3.6. W.A. follows T 4.6. T follows W.A.
3.7. T triggers W.A. 4.7. W.A. triggers T
5. T is same as W.A. 6. T is not same as W.A.
5.1. T and W.A. are similar in 6.1. T differs from W.A. in one way
one way
5.2. T and W.A. are similar in 6.2. T differs from W.A. in several
several respects respects
5.3. T is identical with W.A. 6.3. T is the opposite of W.A.
5.4. T is not different from W.A. 6.4. T is inconsistent with W.A.
|
