An observational study correlating the predicate usage hypothesis of the neurolinguistic programming model and the academic/professional orientation of student subjects

Material Information

An observational study correlating the predicate usage hypothesis of the neurolinguistic programming model and the academic/professional orientation of student subjects
Morreale, Sherwyn P
Publication Date:
Physical Description:
viii, 74 leaves : ; 29 cm


Subjects / Keywords:
Communication -- Cross-cultural studies ( lcsh )
Linguistics, Experimental ( lcsh )
Neurolinguistics ( lcsh )
Communication ( fast )
Linguistics, Experimental ( fast )
Neurolinguistics ( fast )
Cross-cultural studies. ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
Cross-cultural studies ( fast )


Includes bibliographical references (leaves 71-74).
General Note:
Submitted in partial fulfillment of the requirements for the degree of Master of Arts, Department of Communication and Theatre.
Statement of Responsibility:
by Sherwyn P. Morreale.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
18008602 ( OCLC )
LD1190.L48 1985m .M67 ( lcc )

Full Text
Sherwyn P. Morreale
B.A., University ,of Colorado, 1982
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
' Department of Communication and Theatre
l I i .

This thesis for the Master of Arts degree by
Sherwyn P. Morreale
has been approved for the
Department of
Communication and Theatre
Date December 4, 1985

Morreale, Sherwyn P. (M.A., Communication)
An Observational Study Correlating the Predicate Usage Hypothesis of
the Neurolinguistic Programming*^ Model and the Academic/
Professional Orientation of Student Subjects
Thesis directed by Professor Robley D. Rhine
This study investigated differences in language usage
patterns between two dichotomous subgroups in our culture, scien-
tists and humanists. It focused on language differences in terms of
predominant predicate usage patterns (PPUP) and primary representa-
tional systems (PRS). The concepts of PPUP and PRS are derived from
the Neurolinguistic programming theory (NLP)^m of Richard Bandler
and John Grinder (1975, 1976, 1979). NLP refers to the process by
which we use our sensory modalities (visual, auditory, and kines-
thetic) to represent life's events and experiences. The PPUP
hypothesis states that individuals will favor the use of either
visual, auditory, or kinesthetic predicates and this study hypothe-
sized scientists would favor visual predicates and humanists
The study performed content analysis on transcribed inter-
views of 35 right-handed, native-English-speaking graduating
seniors, who had been categorized as scientists (hard majors) or
humanists (soft majors). Using transcribed interviews, two trained
judges, identified the favored use of sensori predi-
cates by subjects. Then the PPUP of the subjects were correlated

with their academic orientation. Audiotapes of the interviews were
analyzed by two NLP-trained practitioners, who also identified for
favored predicate usage by the subjects.
The data generated by both panels of judges (content analysis
vs. NLP-trained) were analyzed utilizing Pearson correlation. Three
of the four judges identified significantly more kinesthetic predi-
cates used by subjects than visual or auditory p<.05. This bias
toward a kinesthetic modality calls into question the PPUP
hypothesis of the NLP theoretical model. When correlating PPUP with
academic orientation, there was significant correlation of kines-
thetic predicate usage with soft majors, but no correlation of
visual predicates with hard majors. This lack of significant cor-
relation did not support the project's hypothesis. As a final ana-
lytical step, within each panel of judges (content-analysis trained
and NLP trained), the two judges were tested in terms of their ten-
dency to similarly identify predicates (V, A, and K). Both panels
demonstrated significant positive correlation regarding predicate
identification, but the content analysis judges indicated a higher
level of inter-rater-reliability than the NLP judges.

My Committee. Robley Rhine, Pam Shockley, Connie Staley. Without
your direction, support, and commitment to academic achievement,
I would not have continued for two and a half years!
My Twenty-Two Volunteers. You who gave of your time and energy on
behalf of my research project, without thought of personal
My Family and Extended Family. You know who you are; and I know
how patiently you've stood by through writing and rewriting.
Thanks to all of you,
Sherry M., M.A.

I. THE PROBLEM................................................ 1
A. Introduction of the Research Problem............. 1
B. General Background to the Study..................... 5
C. Application of Background Information to a
Research Question................................. 8
D. Clarification and Statement of Project
Hypothesis....................................... 10
E. Project Definition................................. 11
1. Delimitations of the Study.................. 11
2. Definition of Terms and Scope of
the Project............................... 12
F. Importance and Justification of the Study.......... 15
II. SURVEY OF THE LITERATURE................................. 24
A. Organization of the Chapter........................ 24
B. Scholarly Literature Review........................ 25
C. Discussion of Relevance of Literature Reviewed
to the Research Problem.......................... 44
III. PROJECT METHODOLOGY....................................... 46
A. Overview of the Chapter............................ 46
B. Project Variables.................................. 46
C. Subject Group Design............................... 47
D. Interview and Field Procedure...................... 50

E. Training of the Judges........................... 52
F. Content Analysis of Transcriptions................. 53
G. Statistical Analysis of Data....................... 54
H. Comparative Analysis of Data....................... 55
I. Limitations and Criticism of the Project
Methodology...................................... 55
IV. PROJECT FINDINGS............................... 57
A. Introduction to the Findings....................... 57
B. Summary of the Findings............................ 58
Tables 1, 2, and 3
V. SUMMARY AND CONCLUSIONS........................... 63
A. Discussion of the Test Results..................... 64
B. Recommendations for Further Study.................. 68
BIBLIOGRAPHY....................................................... 71
BIBLIOGRAPHY OF DISSERTATIONS...................................... 74
(1) Format for Assessment of PPUP in Speech............ 75
(2) Subject Group Design Data.......................... 76
(3) Subject's Questionnaire............................ 77
(4) NLP Predicate List................................. 78
(5) Judges' Report Sheet............................... 81

PREDICATES........................................... 58
(HARD/SOFT)............................................ 59
(Judges 1 and 2) and PANEL II
(Judges 3 and 4)............................. 61
LEVELS OF PANELS I AND II..................... 61

A. Introduction of the Research Problem
What is involved in the interpersonal communication process?
At a most basic level, interpersonal communication involves the
transmission of information through the use of language. Since the
brain, the primary processor and source of communication informa-
tion, is not available for direct observation, language can provide
a source for studying and exploring any differences in human
information processing. Exploring differences in information and
language usage could provide insights that might enhance communica-
Past research and academic writings suggest that although
individuals share a native language, differences in language usage
can impact effective communication. Particularly in the area of
intercultural and intracultural communication, psycholinguists and
researchers have shared a common concern with the relationship of
cultural thought patterns, ways of using language, and effective
communication. Much theory and research suggested that although
individuals share a native language, how they use that language may
vary, and variations in language usage and language patterns may
impact upon interpersonal communication.

One such researcher, Benjamin Whorf (1952), spoke to an
interrelationship of thought patterns and language, and generated
the linguistic relativity hypothesis, popularly called the Whorfian
Hypothesis. This widely accepted hypothesis suggests that language
and especially grammar provide speakers with habitual grooves of
expression. These habitual grooves of expression, or ways of using
the same language, shape the speakers' reality and may vary within a
culture, that is from one cultural subgroup to another. This well
known hypothesis suggests that communication between cultural sub-
groups might be affected by the different ways each subgroup uses
the same language; and communication might be better understood by
studying differences and variations in language usage between
Basil Bernstein (1971), also was concerned with the inter-
relationship of language patterns (or codes), culture, and com-
munication. Bernstein's communication codes theory suggests that
different social structures (or cultural subgroups) within a culture
generate distinctly different linguistic forms of communication
codes. Bernstein theorizes that a cultural subgroup is either an
open or closed social system, using either an elaborated or a
restricted code. According to Bernstein, these distinct and
different codes constrain behavior, transmit culture, and represent
different ways of relating to the world. Like the Whorfian
Hypothesis, Bernstein's codes theory suggests that differences in
linguistic codes within a culture might affect communication between
cultural subgroups.

Another contemporary communication theory, Neurolinguistic
Programming^ (NLP), Bandler and Grinder (1975, 1976, 1979), also
considers differences in human information processing, thought pat-
terns and language patterns, but at an individual level rather than
a global or cultural one. A main tenet of the NLP theory is the
predicate usage hypothesis, which theorizes individual differences
in language patterns in a speaker's choice of predicates in his or
her everyday speech. The predicate usage hypothesis considers
individual differences in language patterns, not cultural or
subcultural differences.
This study focuses on the predicate usage hypothesis, relat-
ing that hypothesis to intra-or-subcultural communication. The goal
of this study is to advance the understanding of subcultural
communication by considering differences in language usage, that is
differences in predicate usage, as related to communication between
cultural subgroups. The study specifically will consider sub-
cultural differences in habitual language patterns that are related
to the predicate usage hypothesis of the NLP theory. That theoreti-
cal hypothesis purports that there are individual differences in
information acguisition and processing -- differences that are
indicated by the predicates habitually used in an individual's
speech -- differences that could affect interpersonal and inter-
cultural communication.
Specifically, the predicate usage hypothesis states that:
(1) Most people who have normal cerebral organization
process visual information using images, auditory

information using sounds, and kinesthetic information
using visceral feelings and bodily sensations; and
(2) the predominant use of visual, auditory or kines-
thetic words and phrases in speech indicates a
habitually favored processing or representational
system. The information provided by that favored
system is used and valued more to cope with and make
sense of the environment. In general, the primary
representational system is the sensory modality that
is favored in assigning meaning to information.
(R. Dilts et al., 1980, p. 1)
SUMMARILY, this hypothesis is saying that:
(1) An individual will be prone to acquiring and processing
more information through one sense than through others; and (2)
that the acquisition and processing differences among the senses
will be reflected in the language patterns, in the verbal symbols,
the words and phrases, that the individual uses habitually.
Taken further and extended, this hypothesis suggests an
effect on communication between speakers of the same language as a
result of differences in individual habitual language patterns.
This extension of the predicate usage hypothesis suggests that
cultural subgroups may have group patterns in information acquisi-
tion and processing that may differ from patterns of other cultural
subgroups and that those subgroup differences may affect commun-
ication among subgroups of a culture. This might extend to
communication between larger cultures.
Such an extension of the predicate usage hypothesis suggests
that a figurative analogy can be drawn from individual interpersonal
communication to communication between cultural subgroups. Drawing
that analogy indicates a potential effect on communication between

speakers of the same language, as a result of subgroup differences
in individual habitual language patterns. For example, one cultural
subgroup may be more dependent upon one sense for information acqui-
sition and processing than the other senses; and another subgroup
may utilize one of the other senses more. Might such differences in
favored sensory modalities and information acquisition and process-
ing affect intercultural communication between these subgroups? If
one subgroup favors one sensory mode and another subgroup favors a
different mode, then the language they use to communicate might
differ also. Since human information processing cannot be observed
but language as a sign of information processing can, then language
patterns provide an accessible place to explore differences in human
information processing. Therefore, exploring and understanding
differences in language usage could further an understanding of
communication between cultural subgroups.
SUMMARILY, this study considers:
(1) differences in information processing and languge usage
patterns of two cultural subgroups, (scientists and humanists); and
(2) the impact of such differences upon the communication process
between those two cultural subgroups.
B. General Background to the Study
What is NLP and what is the predicate usage hypothesis?

1. Individuals receive and represent information about the
environment through sense receptors and organs located throughout
the central nervous system. Each system of sense reception and
transmission to the brain may be called a sensory modality. These
sensory modalities fall into five major categories: vision (sight),
audition (hearing), kinesthesis (body sensations), gustation
(taste), and olfaction (smell). These sensory modalities are
referred to in NLP terminology as representational systems, suggest-
ing that they are systems through which individuals cognitively
represent their impressions of the environment around them. David
Gordon (1978), described representational systems as the sensory
modalities which we have available and use to know (represent) the
world around us. The ways in which we know (represent, experience)
the world are through these sensory portals.
2. As individuals mature, they learn to use or value
information that is provided by one particular representational
system or combination of systems more than information provided by
others. This tendency to use or to value higher one type of sensory
information over others is referred to as representational system
3. Representational System Primacy, according to NLP theory,
is divided into two processes. The two processes are: first, re-
ceiving information and second, assigning meaning to that informa-
tion. These two processes are identified separately. One's lead
system is the sensory modality through which information is brought
to an individual's attention; and one's primary representational

system (PRS) relates to how that information is given meaning.
4. An individual may favor one primary representational
system and a different lead system. That is, an individual may be
very kinesthetic (i.e., make decisions primarily from the way they
feel things), and yet be receiving most of the information that they
are operating off of visually. Conversely, an individual may favor
the same primary and lead systems (i.e., make decisions and receive
information from the same system).
5. A higher valued or favored lead system is identified by
breathing changes, tonal changes, and interjections, but mainly by
certain characteristic eye movements. A favored primary representa-
tional system is identified by body type, size and tension, position
and tonality, but mainly by the predicates used most frequently in
an individual's verbal patterns in everyday language.
6. These identifying verbal language patterns are referred
to as predominant predicate usage patterns (PPUP). Specifically,
the PPUP of an individual will be indicated by the sensory nature or
connotation of the predicates the individual uses most frequently.
The predicates will reflect more strongly one of the major sensory
modalities rather than the other modalities. In this sense, the
identifying words referred to as predicates are the verbs, adverbs,
and adjectives used by an individual.
7. Bandler and Grinder (1975), founders of NLP, defined
predicates as:
. . the portions of a person's experience which
correspond to the processes and relationships in the experi-
ence. These predicates appear as verbs, adverbs, and adjec-
tives in the sentences which the client uses to describe his

experience, (p. 9)
8. NLP expert, Robert Dilts (1980), of the Neurolinguistic
Programming Behavioral Technology Institute, conceptualized predi-
cates thusly:
. . the predicates an individual uses to describe the
experience will indicate the way in which they perceive and
make sense of that experience, (p. 1)
SUMMARILY, Neurolinguistic Programming^111 (NLP) and the predicate
usage hypothesis (PPUP) suggest that there are individual differ-
ences in language patterns, that is, in the predicates used by an
individual to describe his or her experiences.
C. Application of the Background Information
to a Research Question
How does the predicate usage hypothesis apply to communica-
tion between cultural subgroups?
The predicate usage hypothesis suggests some possible effects
on the exchange of information and communication between individuals
and between cultural subgroups.
I. Regarding the effect of predominant predicate usage patterns on
communication between individuals, one may hypothesize that:
1. Individuals communicate relatively well with other
individuals who favor the same PRS and PPUP.
2. If an individual who has one favored PRS and PPUP
attempts to communicate with someone who has a different favored PRS

and PPUP, there will be relatively poor communication.
3. If individuals can be educated to recognize differences
in PRS and PPUP, then by adapting their PPUP to the PPUP of the
other individual, those individuals may improve their communication
As suggested earlier in this introduction, a figurative
analogy can be drawn from individual interpersonal communication to
intercultural communication between subgroups. If a significant
number of individuals in one cultural subgroup tend to favor one
particular PRS and PPUP, and individuals in another cultural
subgroup tend to favor a different PRS and PPUP, then the exchange
of information and communication interactions between the two
subgroups will be affected.
II. Regarding the effect of predominant predicate usage patterns on
communication between cultural subgroups, one may hypothesize that:
1. Individuals within one subgroup will communicate
relatively well with fellow subgroup members who exhibit a similar
2. There will be relatively poor communication between
individuals of different subgroups, who exhibit a different PRS and
3. Communication could be improved between members of
different subgroups through education regarding PRS and PPUP differ-
ences and through the adaptation and modification of their verbal
patterns to accommodate their differences in PRS and PPUP.

SUMMARILY, application of the predicate usage hypothesis suggests
that there may be cultural subgroup differences in language patterns
(in the predicates used by varying subgroups) that could affect
communication between the subgroups.
D. Clarification and Statement of Project Hypothesis
This project hypothesizes that differences in primary repre-
sentational system (PRS) and in predominant predicate usage patterns
(PPUP) between cultural subgroups may affect the exchange of
information and effective communication between cultural subgroups.
The project studies the language patterns of two major cul-
tural subgroups within the American culture. It explores differ-
ences in predominant predicate usage patterns between two academic/
professional subgroups; individuals of a hard (scientific) orienta-
tion (hard majors), and individuals of a soft (humanities) orien-
tation (soft majors). (See p. 14)
If significant differences in predicate usage (language pat-
terns) exist between the two subgroups, then effective communication
between these two subgroups could be affected.
The Following Research Hypothesis Will
be Explored Empirically:
1. Members of American culture subgroup 1 (ACS-1) (scien-
tific oriented individuals) (hard majors) will exhibit significantly
different PPUP from members of American culture subgroup 2 (ACS-2)
(humanities-oriented individuals) (soft majors).

The Following Hypotheses will be Suggested
but will not be Explored Empirically;
1. The significant differences in PPUP between the two cul-
tural subgroups suggest relatively good communication interactions
between the individuals of the same subgroup but relatively poor
communication interactions between individuals of differing sub-
2. Communication could be improved between the two cultural
subgroups through remedial education regarding differences in PPUP
and adaptation and modification of PPUP to accommodate the differ-
ences .
E. Project Definition
1. Delimitations of the Study;
The original intention of this project is neither to support
nor negate the predicate usage hypothesis. Rather it intends to
explore intracultural language usage differences between two
cultural subgroups: hard majors or individuals of a hard science
academic/ professional orientation and soft majors or individuals of
a humanities orientation, (Katz and Salt, 1981). The predicate
usage hypothesis has been explored in other studies which are
described in the literature review. The purposes of those studies
were to test the predicate usage hypothesis in terms of its ability
to generate treatment and corrective techniques that would establish
rapport and improve communication. Such corrective techniques
involve matching the predominant predicate usage patterns of other

speakers in order to establish rapport and treat or prevent communi-
cation problems or disorders. The correction of communication
problems between cultural subgroups using the predicate matching
techniques, is not the subject matter of this study. Rather, this
study is delimited to the task of ascertaining the existence of any
differences in predicate usage patterns that may suggest the future
use of corrective predicate matching techniques. Any study or
measurement of corrective techniques to overcome language usage
differences between cultural subgroups is outside the scope of this
SUMMARILY, this study is limited to exploring intracultural language
usage differences, specifically differences in predominant predicate
usage patterns between hard major individuals of a hard science
academic/professional orientation and soft majors or individuals of
a humanities orientation. It will not explore techniques for
improving communication between hard and soft majors by using the
predicate usage hypothesis or any other corrective technique.
2. Definition of Terms and Scope
of the Project:
Delimitations are further understood by defining and clarify-
ing the key terms involved in the hypothesis:
Neurolinguistic Programming^ (NLP): A model of communica-
tion proposed by Bandler and Grinder (1975, 1976, 1979), which is
based upon perceptual and linguistic analysis of the represen-
tational systems by which individuals receive and take in informa-

tion and assign meaning to that information. The design of the NLP
model was based on the systematic observation of communication
patterns, including habitual language patterns. NLP theorizes that
most people who have normal cerebral organization process visual
information using images, auditory information using sounds, and
kinesthetic information using feelings and bodily sensations.
According to the NLP model, the predominant use of visual, auditory,
or kinesthetic predicates in a person's habitual language patterns,
indicates which sensory system is in use and is most favored by that
person. The NLP model does not focus on olfactory and gustatory
predicates because those two sensory modalities are inconsequential
in our culture. Predominant Predicate Usage Pattern Hypothesis
(PPUP): A tenet of the NLP model of communication, (Bandler and
Grinder, 1975, 1976, 1979), which theorizes individual differences
in a person's choice of predicates in his or her habitual language
patterns. This tenet was based upon the analysis of communication
interactions in which favored or predominant use of sensorily
connotative predicates was observed. According to the PPUP
hypothesis, persons will demonstrate the use of a favored or
predominant sensory system through the predominant use of either
visual, auditory, or kinesthetic predicates.
Predicates: According to NLP theory, (Bandler and Grinder,
1975, 1979):
predicates appear as verbs, adverbs, and adjectives in a
sentence which a client uses to describe his or her experi-
ence. The predicates an individual uses to describe an
experience will indicate the way in which they perceive and
make sense of that experience, (p. 9)

According to the Harbrace College Handbook (1972), the predi-
cate is defined as:
... a basic grammatical division of a sentence. A
predicate is the part of the sentence compromising what is
said about the subject. The complete predicate consists of
the main verb along with its auxiliaries (the simple predi-
cates) and complements and modifiers. Ex: He runs through
the house. Runs is the simple predicate, runs through the
house is the complete predicate, (p. 77)
Habitual Language Patterns, (as opposed to linguistic
style): Linguistic style is considered a "characteristic way of
making non-semantic linguistic choices which consistently distin-
guish among different comparable users of language," (Sandell,
1977). Impacting stylistic variables that are often researched
are: considerations of vocabulary variability, the use of excep-
tional words and key words, or types of words, use of particular
parts of speech idiosyncratically, and differences in structural
organization. A characteristic or habitual language pattern is
considered present "when a certain linguistic style choice is
consistently repeated," (Sandell, 1977).
The habitual language pattern considered in this study
involves consistent and predominant use of either visual, auditory,
or kinesthetic predicates.
Academic/Professional Orientation: (Hard vs. Soft Major).
A hard (scientific) academic orientation includes the science
subjects, specifically mathematics, biology, engineering, economics,
physics, and chemistry.
A soft (humanities) academic orientation includes the humani-
ties subjects, specifically English, history, theatre, foreign

language, religion, and philosophy. (More extensive explanation of
the operational definitions of hard and soft major is presented
under subject group design in the methodology section).
F. Importance and Justification of the Study
Addressing a communication gulf between cultural subgroups is
the task, but, as earlier delimited, this project addresses only one
facet of communication between cultural subgroups, i.e., language
differences between scientists and humanists. By understanding such
language differences, it is hoped that some new bridges of under-
standing may be initiated between the two groups.
Justification and support for investigating this communica-
tion gulf are derived from a variety of sources: some philo-
sophical; some practical in a contemporary business sense; and some
empirical in nature (that is, similar empirical research efforts).
The importance of intracultural differences in vocabulary
between speakers of the same language was stated by psycholinguistic
researcher Edith Folb (1982):
Culture and language are inseparably intertwined. What
we think about and how we think about it are direct func-
tions of our language. However, when two groups of people
even within the same culture share a language, this does not
necessarily mean that vocabulary is used to characterize
similar or identical experiences. One of the greatest
barriers to effective communication, whether it be intra-
cultural or intercultural, is the lack of shared experience
and a common vocabulary to define them. (p. 241) ,
Initial philosophical justification for this project is pro-
vided by the writings of historian and theorist, C. P. Snow (1963).
Snow specifically referenced the intercultural communication problem

between scientists and humanists in Two cultures, A Second Look
(1963). Although Snow did not conduct empirical research, he did
point to an intercultural barrier of some moment between his
scientifically-oriented and his literary-oriented colleagues. He .
referred to separate cultural subgroups, the scientific and the
non-scientific or humanistic, and described the polarization of the
two as sheer loss to us all, practical, intellectual, and creative
loss" (p. 11).
Snow concernedly stated that there seemed to be no place
where the two cultures met. He felt that:
. . the clashing point of the two disciplines, the two
cultures, ought to produce creative changes. But the two
cultures can't talk to each other. This inability of the
two cultural subgroups to talk and to communicate is a
universal phenomenon, a cultural divide all over the western
world. The separation between the scientists and non-scien-
tists is much less bridgeable now than it was 30 years ago.
It may well be that this process has gone too far to be
reversible. But it is a disastrous process for the purpose
of a living culture, (p. 22)
Not only did Snow discuss the existence of a gulf between
scientists and humanists, he also clarified the impact of that gulf
and communication barrier on our society and culture:
. . the success of industrialization and the Scientific
Revolution is dependent on minimizing polarization of the
two cultures ... we have not been as quick as we should to
draw the right consequences of the Revolution, very largely
because of the division of the cultures. It has been hard
for politicians and administrators to grasp the practical
truth of what scientists were telling them. (p. 80)
In a final concerned statement, Snow asserts that:
.it is dangerous to have two cultures which can't or
don't communicate. In a time when science is determining
much of our destiny, that is, whether we live or die, it is
dangerous in the most practical terms! (p. 98)

This "dangerous" communication problem described by
C. P. Snow, justifies research efforts that would explore this pro-
blem empirically.
More contemporary justification for this project is provided
by the writings of Ben Shneiderman (1980), an author in the field of
human factors analysis and design, who discussed another aspect of
the differences between scientists and humanists. In Software
Psychology: Human Factors in Computer and Information Systems
(1980), Shneiderman was concerned with the practical and psycho-
logical aspects of human performance and variations in cognitive
style. He defined cognitive style as:
... a mode of information processing, the two poles
being analytic (scientific) and heuristic (humanistic).
Analytic implies sequential, linear, verbal symbolic
processing, or left-oriented, and heuristic implies intui-
tive, global, pictorial processing, or right-brain oriented.
(p. 56)
Shneiderman was concerned that unidentified and misunderstood
differences in cognitive style, that is differences of an analytic
vs. heuristic nature, could impact upon human performance in the
information and computer field.
. . difficulty in implementing management information
systems can be the result of cognitive style mismatches
between users and information design. Behavioral research
in human use of computers, though limited and new, is making
a place for itself. As awesome and complex as the computer
may seem, many individuals learn to effectively use this
tool. The difference between those who approach the
computer and those who avoid it is poorly understood.
(p. 56)
This area of concern, that is differences in analytic or
heuristic cognitive, style, cannot be easily observed but differences
in language usage can be. This concern for intangible differences

in cognitive style provides further justification for exploring
tangible language differences between scientists and humanists.
In addition to the perspectives of C. P. Snow and Ben
Shneiderman, further justification was provided earlier in this
introduction by citing the concepts related to culture and language
of Benjamin Whorf (1952), and Basil Bernstein (1971). Both Whorf
and Bernstein were concerned with the theoretical relationship of
cultural thought patterns and differences in language usage. The
Whorfian linguistic relativity hypothesis and Bernstein's communi-
cation-codes theory suggest differences in language usage between
cultural subgroups within a culture. Both theories, Whorf's and
Bernstein's, imply a possible effect on communication between sub-
groups affected by language usage differences. Thus this study of
subgroup differences is further justified.
C. P. Snow, Shneiderman, Whorf, and Bernstein all point to
and share a common thesis, that language usage differences and cul-
tural thought patterns are inextricably interrelated. This project
intends to investigate empirically a facet of that interrelationship
using NLP and the predominant predicate usage hypothesis as the
vehicle or variable of investigation. The communication problem
between scientists and humanists has been philosophically discussed
by Snow and Shneiderman; and Whorf and Bernstein have pointed to
subgroup language usage differences. This project is hypothesizing
that the subcultural communication problem cited by Snow and
Shneiderman may be related to the subcultural language usage differ-
ences described by Whorf and Bernstein.

Empirical support is also necessary to properly justify this
project. Such empirical justification is provided by research
efforts which have generated significant findings in similar areas
of inquiry. Several of these related research efforts will be
described to further justify this project. More qualitative and
quantitative descriptions of other related studies are included in
the literature review.
In a similar language study Katz and Salt (1981), also cor-
related parts of speech and academic orientation. That study pre-
sented evidence of differences in language use by adults which
seemed related to cerebral hemispheric activity and academic
orientation. The verbal language patterns of academic subgroups
indicated that left-eye-movers (soft majors) used a significantly
greater number (p<.10) of adjectives than right-eye-movers (hard
majors), and hard majors used more nouns than did soft majors. This
study presented experimental evidence that the direction of eye
movements (indicating cerebral hemispheric activity), and academic
orientation are related to individual stylistic differences in
language usage. Like the study of Katz and Salt, the project
presented here will investigate the language patterns of hard and
soft majors, but predicate usage patterns will be explored rather
than nouns and adjectives. By way of empirical justification, if
hard majors and soft majors differ in noun/adjective language
patterns, it is hypothesized that differences in predicate usage
also may be found.
In another published study regarding stylistic differences in

the style of children's language, (Staley, 1982), a content analysis
system was developed that initially included two major language
categories: (1) Descriptive language was categorized stylistically
as empirical, non-judgmental, non-interpretive, and non-emotional;
descriptive language is language that relates to observable visual
features of a stimulus. (2) Interpretive-emotive language was
categorized as creative, personal language, or language that des-
cribes a psychological (or kinesthetic) relation to a stimulus.
That study found significant differences in language use in young
children, with 4-year-old males using more descriptive language and
4-year-old females more interpretive-emotive language. The differ-
ences lessened as the children matured with differences in the use
of descriptive language still favoring males at the 8-year-old level
but favoring females at the 16-year-old level. The system of
language categorization used in that study is similar to the
categorization system of the project being justified. Descriptive
language might be considered similar to a visual orientation and
would be expected of hard majors. Interpretive-emotive language
might be considered similar to a kinesthetic orientation and would
be expected of soft majors. Since the hard majors (science and
engineering) are somewhat male dominated, and the soft majors
(humanities) are somewhat female dominated, justification for this
study of predicate usage is suggested.
Further justification for this project is provided by another
pilot study, (Morreale, 1983). In that study, the verbal language
patterns of 32 college students, 14 hard majors and 18 soft majors,

were analyzed for differences in predicate usage. The hard majors
were identified by their significantly greater use of visual
predicates and the soft majors by their significantly greater use of
kinesthetic predicates, at a p<.05 level of significance. Based on
identifying the predominant use of one particular type of predicate,
the following PPUP identifications were made by an NLP judge: of
the 14 hard majors, 10 were identified as visual, 2 auditory, and 2
kinesthetic, and of the 18 soft majors, 15 were identified as kines-
thetic, 2 auditory, and 1 visual. Although the control of impacting
variables in this pilot study was problematic, the results do
suggest justification for further study.
These three empirical studies, Katz and Salt, C. Staley, and
the Morreale predicate usage pilot project all suggest the possi-
bility of correlation among language usage patterns, cerebral hemis-
pheric activity, and academic orientation. The project being justi-
fied explores these correlations, focusing particularly on predicate
usage as the language pattern being investigated.
Final justification regarding the importance of this study is
provided by a published article entitled, "Assessment of Primary
Representational Systems with NLP: Examination of Preliminary
Literature," (Dorn et al., 1983). That article reviews findings in
academic literature that have addressed the identification of PRS.
It summarizes:
. . the inconsistent findings of studies on Neurolin-
guistic Programming and recommends some areas that need to
be examined if various NLP claims are to be verified.
(p. 161)
According to Dorn et al., after reviewing the bulk of NLP-

related research, some methodological shortcomings are in need of
. . the research on NLP and the identification of pri-
mary representational systems so far has been disappoint-
ing. The most disturbing factor reappearing throughout NLP
research is the lack of a reliable method of assessing PRS.
It is difficult to overestimate the importance of such cri-
terion. Existing research has been valuable in that it
provides directions for us to pursue and points to method-
ological concerns that must be addressed. But, if various
NLP strategies are proposed to affect individuals in dif-
ferent ways and are to be effective interventions for
change, then the PRS of the individual must be responsibly
diagnosed, (p. 166)
Of concern to Dorn et al., is NLP's major thrust on training
rather than methodological research. NLP-trained practitioners are
certified as capable of identifying an individual's PRS based on an
intuitive listening process. Dorn et al., found this method of
assessing PRS lacking in terms of scientific reliability. The cur-
rent study endeavors to use a more reliable method for assessing
PRS. This study uses a content analysis process for identifying the
predicates used by each subject. By contrast to the PRS assessment
methods criticized by Dorn et al., this study responds to those
criticisms, by utilizing content analysis methodology. Thus, in
addition to investigating the primary hypothesis regarding subcul-
tural differences, the current study also provides a different
methodological model for PRS assessment.
In summary, this study is validated as a research effort by
the varying philosophical and empirical justifications set forth.
The gulf between scientists and humanists ultimately may be bridged
by this and similar empirical research efforts.

The following chapters will provide:
(1) a review of academic literature related to the project;
(2) the field methodology utilized by the project;
(3) results of the field research and discussion of the
test findings; and
(4) recommendations for further study based upon the

A. Organization of the Chapter
The investigation of academic literature for this project
reviews published and unpublished works (books, dissertations and
theses), related to the research problem set forth. Three areas of
academic inquiry are explored and overviewed:
Communication Between Scientists and
Humanists and Subcultural Differences
in Cognitive Style:
These writings provide a theoretical and philosophical frame-
work for this project. (The work of C. P. Snow and other similar
writers is reviewed.)
Language Theory and
Language Patterns Studies:
These writings include the ideas of linguistic theorists,
Sapir/Whorf and B. Bernstein. (Empirical research work on language
patterns also is reviewed.)
The NLP Theory and the Predicate
Usage Hypothesis:
These writings consider the nature and quality of empirical
research that has been conducted. (Evaluative judgments are made of

the research.)
Following each of the three areas of inquiry, there is a
discussion of the relevance of that literature to the research pro-
blem. At the end of the chapter there is commentary on the entire
survey of literature.
B. Scholarly Literature Review
Communication Between Scientists
and Humanists and Cognitive
Style Differences:
Although concern for communication between scientists and
humanists was expressed twenty years ago by C. P. Snow (1963),
little empirical research has focused on that concern. However,
some research has investigated cognitive style differences between
similar subgroups. The writings of C. P. Snow are reviewed as a
philosophical base for this project, along with several writings
focusing on cognitive style differences between cultural subgroups.
C. P. Snow (1963), expressed considerable philosophical
concern for the problem of communication effectiveness between
scientists and humanists. He described his intuitive awareness of
the existence of communication barriers between his scientific and
literary colleagues. Snow said:
... I felt I was moving among two groups -- comparable
in intelligence, identical in race, not grossly different in
social origin, earning about the same incomes -- who had
almost ceased to communicate at all, who in intellectual,
moral, psychological climate had so little in common. The
intellectual life of the whole western society is increas-
ingly being split into two polar groups . literary
intellectuals (humanists) at one pole -- at the other,
scientists -- and between the two a gulf of mutual incom-
prehension, sometimes hostility and dis'like, but most of all

lack of understanding, (p. 2)
Snow pointed to potential negative ramifications of this gulf
and lack of understanding between scientists and humanists. He
stated that the full success of industrialization and scientific
revolution is dependent on minimizing the polarization of the two
cultures. He felt that the spread of the benefits of scientific
revolution have been negatively impacted by polarization, and that,
"we have not been as quick as we should to draw the right conse-
quences, very largely because of the division of the cultures. ."
(p. 80).
Snow also discussed the communication gulf between scientists
and humanists as a function of differences in human information
processing. He referred to "different kinds of mental activity,"
(p. 22), i.e., cognitive style differences between the two groups.
He stated that:
the reasons for the existence of the two cultures are
many, deep, and complex, some rooted in social histories,
some in personal histories, and some in the inner dynamics
of the different kinds of mental activity themselves.
(p. 22)
Snow's suggestion of differences in mental activity between
cultural subgroups is supported by other writings in academic
literature. Other researchers have labeled mental activity differ-
ences as differences in cognitive style and have observed those
differences individually and subculturally.
Cognitive style has been conceptualized by researchers,
Gonzales and Roll (1985), as:
the particular way in which individuals perceive,
organize and label stimulus in order to arrive at a basis of

similarity among stimuli. This organizing and labeling,
whether overtly spoken or' not, is said to be the essential
process of cognitive style, (p. 190)
Gonzales and Roll cite research suggesting that not only will
individuals be consistent in how they organize and label stimuli,
but that the same individuals also will show stable individual
preferences in modes of organizing a variety of stimuli. These
individual preferences and differences in cognitive style support a
thesis that cognitive style difference also could exist between
groups of people, that is between cultural subgroups.
Academic interest in cognitive style has led to empirical
research on variations in human performance on computers that is
related to cognitive style differences. Shneiderman (1980), defined
cognitive style as "a mode of information processing that is polar,
the two poles being analytic and heuristic" (p. 56). Analytic
implies sequential, linear, verbal symbolic processing, or left-
brain oriented, and heuristic means intuitive, global, pictorial
processing, or right-brain oriented.
Shneiderman was studying differences in cognitive style as
related to human factors design in computers and information sy-
stems. His research efforts suggest that difficulty in implementing
information systems can be the result of cognitive mismatches
between the users of the system and the information design.
Shneidermans interest in the more cognitive or psychological
aspects of human performance variation has focused on behavioral
research in human use of computers, a new and somewhat limited field
of study. According to Shneiderman the difference between those who

approach the computer and those who avoid it may be a polar differ-
ence, analytic as opposed to heuristic.
Subgroup differences in cognitive style, between those who
approach computers comfortably and those who do not, support the
thesis of cognitive style differences between analytic (scientists)
and heuristics (humanists). Shneiderman suggests further empirical
investigation of the cognitive style differences between these two
subgroups, considering a thesis that one group may be cognitively
analytic and the other cognitively heuristic. Such a difference in
cognitive style could effect communication within, as well as be-
tween, these two groups.
The nature of differences in cognitive style (i.e., analyti-
cal vs. heuristic, left-brained vs. right-brained, or scientist
vs. humanist) was clarified by Joseph Bogen (1977). Bogen compiled
a list of terms used by psychologists who, over many years, have
been groping toward a definition of left-brain and right-brain
styles of knowing. Here is a slightly truncated version of Bogen's
list of terms:
LEFT-BRAIN (Analytic) RIGHT-BRAIN (Heuristic)
Intellect Intuition
Convergent Divergent
Intellectual Sensuous
Deductive Imaginative
Active Receptive
Discrete Continuous
Realistic Impulsive
Propositional Inductive
Transformational Associative
Lineal Nonlineal
Historical Timeless
Explicit Tacit
Objective Subjective

This list describes possible differences in cognitive style
between the analytically-oriented and heuristically-oriented, that
is between scientific as opposed to humanistic individuals. However
it only describes such differences superficially, and does not fully
support such a cognitive style dichotomy.
In summary of this section, several writings have been pre-
sented that point to a problem in cognitive style differences
between cultural subgroups.
C. P. Snow identified a communication gulf between scientists
and humanists as "sheer loss to us all, practical, intellectual, and
creative loss" (p. 11). Snow did not conduct empirical research
into the nature of the cognitive style differences between scien-
tists and humanists. Rather he called attention to a communication
problem and to its practical political ramifications. Snow did
suggest that differences in mental activities and cognitive style
might be related to the gulf between scientists and humanists, and
later writers supported that notion.
Gonzales and Roll (1985), summarized a body of writers and
theorists who had observed individual differences in cognitive
style. Shneiderman (1980), identified a polarity in cognitive style
that he termed analytic at one pole and heuristic at the other.
Bogen's summary of psychological terms associated with differences
in cognitive style also described an analytical/heuristic polarity,
possibly of a subcultural nature.
Other theorists and writers have linked differences in cogni-
tive style to subcultural differences in language usage. This

review now cites several of those linguistic theories and studies.
Language Theory and Language
Patterns Studies:
The inter-relationship of cognitive style (thought patterns),
language patterns, and culture has been considered theoretically as
well as by empirical investigation. The writings of linguistic
theorists, E. Sapir, B. Whorf, and B. Bernstein are reviewed, along
with several empirical studies focusing on subcultural language
Edward Sapir and Benjamin Whorf (1931 and 1952), generated
the linguistic relativity hypothesis, now popularly called the
Whorfian hypothesis. The central idea of this hypothesis is that
language functions not simply as a device for reporting experience,
but also as a way of defining or shaping experience for its speak-
ers. Whorf defined this hypothesis (1952), stating that the
linguistic system (the words and grammar) is not merely a reproduc-
ing instrument for voicing ideas, but rather is itself, the shaper
of ideas, the guide for the individual's mental activity. Although
the Whorfian hypothesis has not been proven or disproven, it does
suggest that language, particularly words and grammar, plays a large
and significant role in cultures (and subcultures). Far from being
simply a communication mechanism, language is itself a shaper of
perceptions for speakers in a culture. Language, and especially
grammar, provide speakers with habitual ways of perceiving reality.
The linguistic rrelativity hypothesis further suggests
(Hoijer, 1954), that if language (language patterns) differ markedly

from each other, significant barriers to intercultural communication
may exist. In an analysis of the Whorfian hypothesis, Hoijer
observed that these cultural communication barriers take on even
greater importance when it is realized that the phenomenon of
language is largely outside the critical consciousness and control
of the speaker. Hoijer's analysis of the linguistic relativity
hypothesis related the hypothesis' theory to intercultural and
subcultural communication.
Further theoretical analysis of the inter-relationship of
linguistic forms (or codes), language, and culture was provided by
Basil Bernstein (1971). Bernstein's communication-codes theory
suggests that different social structures, or subgroups, within a
culture, generate distinctly different linguistic forms, or codes,
of their own. The purposes of these distinct and different sub-
cultural codes are to constrain behavior, to transmit culture, and
to represent ways of relating to the world. Like the Whorfian
hypothesis, Bernstein's theory of subcultural communication codes
suggests possible communication barriers as a result of differences
in linguistic codes between cultural subgroups.
Edith Folb (1982), empirically studied the relationship of
subcultural differences in language usage and communication between
subgroups. Folb, in a study of inter-racial language differences,
suggested that subcultural vocabulary (word choice and language
usage) differences are worth exploring. She stated that vocabulary
is the part of language that provides the most accessible place to
begin exploration of shared and disparate experiences. In Folb's

study a potential for miscommunication became apparent, as a result
of the significant number of phonetically identical items that were
assigned different meaning by black as opposed to white youths.
Samovar and Sanders (1982), in a study of language patterns
of a deviant subculture, also observed and commented on subcultural
language usage differences. According to Samovar and Sanders:
One way to gain insight into any subculture is to examine
its use of language. Subcultures can be examined in terms
of a combination of language, values, and behavior. This
method of analyzing subcultures suggests that experience and
language work in tandem, and it is often difficult to
determine which is the voice and which is the echo.
(p. 249)
The Samovar and Sanders' study analyzed the language patterns
of prostitutes as a deviant subculture. The study focused on pros-
titutes' idiosyncratic use of language to accomplish client
manipulation and to promote group solidarity. As a subcultural
group, prostitudes demonstrated significant idiosyncratic language
usage patterns. Such evidence of subcultural language patterns in
one subgroup suggests the existence of similar patterns in other
subcultural groups.
Subcultural differences in language usage also were studied
empirically by Katz and Salt (1981), (language patterns of hard and
soft majors); and by C. Staley (1982), (language patterns of
children compared by sex).
The Katz and Salt study (1981), correlated subjects' language
usage, their patterns of eye-movements, and their academic orienta-
tion. The study presented evidence of differences in language usage
which seemed related to cerebral hemispheric activity and academic

orientation. The project analyzed the eye-movements and language
patterns of 25 right-handed native English speaking college students
between 18 and 30 years of age. Subjects were divided into two
groups based on the direction of the majority of their eye-move-
ments. Observation of eye-movements indicated that those subjects
whose eyes moved more often to the right were hard majors (of a
scientific academic orientation). Subjects whose eyes moved more
often to the left were soft majors (of a humanistic academic
orientation). Right eye-movers expressed a preference for science
courses and left eye-movers for humanities courses. Content
analysis of the verbal language patterns of the two academic
subgroups (hard and soft majors) indicated that left-eye-movers
(soft majors) used a significantly greater number of. adjectives than
right-eye-movers (hard majors), and right-eye-movers (hard majors)
used more nouns than soft majors. This study empirically supported
the relationship of three variables: eye-movement direction
(indicating cerebral hemispheric activity and cognitive style);
academic orientation (hard or soft major); and language usage
differences (nouns vs. adjectives).
The study of C. Staley (1982), considered stylistic differ-
ences in the style of children's language. The pilot study for this
project developed a content analysis system that initially included
two major language categories. The first category, descriptive
language, was described stylistically as empirical, non-judgmental,
non-interpretive, and non-emotional; descriptive language is
language that relates to observable features of a stimulus. A

second category, interpret!'ve-emotive language, was described
stylistically as creative, personal language, or language that
describes a psychological relation to a stimulus. The study found
significant differences by sex in language use of young children.
Four-year-old males used more descriptive language and four-year-old
females used more interpretive-emotive language. The differences
lessened as the children matured with descriptive language use still
favoring males at the eight-year-old level, but switching to
favoring females at the sixteen-year-old level. Relating Staley's
system of language categorization to predicate usage patterns,
descriptive language might be likened to a NLP visual orientation
and interpretive-emotive to a NLP kinesthetic orientation.
The studies of Katz and Salt and C. Staley suggest further
investigation of academic orientation and of varying language usage
patterns. Katz and Salt found evidence of language usage differ-
ences between hard and soft majors in terms of nouns and adjec-
tives. C. Staley developed another categorization system of
language usage differences. If the Staley categorization system
were related to hard and soft majors, it might reveal descriptive
vs. interpretiveemotive differences between academic subgroups. In
any case, both studies support the reality of subcultural
differences in language use.
A pilot study, (Morreale, 1983), also correlated academic
orientation and a categorization system of language usage patterns.
In that study, the oral verbal language patterns of 32 college stu-
dents, 14 hard majors and 18 soft majors, were analyzed. Audiotapes

of subjects' interviews were analyzed by an NLP practitioner who
listened to the tapes for predominant predicate usage patterns. The
categorization system used was based on the predominate predicate
usage hypothesis and included visual, auditory, and kinesthetic
categories. The hard majors were identified by their significantly
greater use of visual predicates, and the soft majors by their
significantly greater use of kinesthetic predicates, at a p<.05
In summary of this section, several theoretical and several
empirical perspectives have been presented. Both the Whorfian hypo-
thesis and Bernstein's codes theory support an inter-relationship of
thought, language, and culture or subculture. The empirical
investigations described also suggest cultural and subcultural
differences in language usage. Folb's study identified vocabulary
differences between black and white youths. Samovar and Sanders
identified idiosyncratic language usage patterns of prostitutes as a
deviant subculture. Katz and Salt, and Morreale, identified
differences in language usage (nouns vs. adjectives; visual audi-
tory, or kinesthetic predicates) between academic subgroups.
C. Staley identified language usage differences between young males
and young females.
A statement (Asuncion-Lande 1983), from the Intercultural
Communication Theory Annual, summarizes the importance of language
investigations in relation to intercultural understanding.
The potential of language theory for building a theory in
intercultural communication is quite significant. It may
provide clues for developing the linkage between language
and communication and communication and culture, (p. 253)

NLP Theory and the Predicate Usage Hypothesis
A language theory, perhaps applicable to intercultural com-
munication, was originally introduced as a therapeutic counseling
tool. In the mid-1970's this therapeutic model of communication,
based on transformational linguistics, was proposed by Bandler and
Grinder (1975, 1979), Grinder and Bandler (1976), entitled Neurolin-
guistic Programming^01 (NLP). Since that time, NLP has been the
subject of several more books (Dilts, Grinder, Bandler, Bandler, and
Delozier, 1980; Lankton, 1980 ; Cameron-Bandler, 1985). As a
communication tool, NLP has been presented to many thousands of
people in professional workshops, (Harmon & O'Neill, 1981). As a new
theoretical model of communication and counseling, it has received
considerable research attention, to be described herein.
The NLP Model, as proposed by Bandler and Grinder, makes
several assumptions about cognitive style and the observable cues
related to cognitive style. NLP assumes that people: (a) organize
experience into representational systems reflecting one or more
sensory modes (visual, auditory, or kinesthetic); (b) vary in
tendencies to encode experience primarily in one sensory mode or
another; (c) express such individual differences through choice of
verbal predicates (I see . I hear . I feel . .) and
through eye-movements (see Falzett, 1981); and (d) communicate
best with others who use the same representational or predicate
Research attempts to examine these assumptions have centered
on two points: (1) differences in verbal predicates and eye-move-

merits and (2) communicating better using similar predicate
systems. The majority of published NLP research has focused on
testing these two more novel components of the NLP model: eye-move-
ments, (Dorn et al., 1983; Falzett, 1981; Owens, 1977; Shaw, 1977;
Thomason etal., 1980), and on predominant predicate usage, (Dorn,
1980; Dowd and Pety, 1982; Falzett, 1981; Gumm et al., 1982; Owens,
1977; Shaw, 1977; Yapko, 1981). NLP proponents, Dilts et al.,
(1980), summarily proposed that the observable cues, eye-movements
and predicate usage patterns correspond with the favored cognitive
style or manner which people use to organize their subjective
The developers of NLP referred to this favored cognitive
style or preference for one sensory mode over others as the primary
representational system (PRS), (Bandler & Grinder, 1975). They
outlined two specific methods for identifying PRS. One is the
predicate usage method and the other is the eye-movement procedure.
This review considers preliminary findings in published and unpub-
lished literature that have addressed the identification of PRS
through predicate usage.
Published Research on Predicate Usage
(Books and Academic Articles)
Research findings on identifying PRS through predicate usage
have been inconsistent and inconclusive at best. A surprising
aspect of predicate usage research, thus far conducted, is that it
primarily has been of an outcome nature. What mainly has been
researched is the outcome or effect of matching predicates on

variables such as trust, rapport, etc. Such outcome research
assumes that the predicates or PRS of subjects are accurately
determined or assessed. Only two studies, (Falzett, 1981; Gumm et
al., 1982), called into question, the ability to accurately assess
predicates, before proceeding with outcome or effects research.
Shaw (1977), in one of the first NLP studies in this area,
found that participants whose PRS was identified through the predi-
cate usage method did not respond any differently to a similar PRS
than did those who were exposed to a dissimilar one. It should be
noted that participants in this study were presented visual, audi-
tory, or kinesthetic forms of the same videotaped story and were
identified as auditory or kinesthetic (no participant was identified
as possessing a visual PRS).
Dorn (1980), compared the levels of attraction held for coun-
selors when they were matched with clients on the basis of preferred
predicate usage. Participants' verbal responses to a number of
images were tape recorded, and trained raters determined each par-
ticipants' PRS. These participants then listened to three audio-
tapes, each containing a series of predicates (either visual, audi-
tory, or kinesthetic) that were spoken by a PhD-level counseling
psychologist. Participants were asked to indicate their degree of
social attraction for each counselor on the basis of preferred PRS.
The hypothesis that clients would be more attracted socially to a
counselor with a similar PRS was not supported. Clients who had a
visual PRS, however, did rate all counselors as more attractive than
did those with a kinesthetic PRS. It should be noted that after

interviewing 130 participants, only 3 were identified as having an
auditory PRS; thus, only participants with either a visual or kines-
thetic PRS were included in the data analysis.
Two additional studies, using a design similar to that used
by Dorn (1980), matched counselors and clients on the basis of
predicate usage. In the first study, Down and Hingst (1981),
investigated the effects of predicate matching, predicate mismatch-
ing, and no matching of predicates, on the perception of counselor
social influence and the counseling relationship with clients, in an
initial interview. After several training sessions, counselors were
instructed to consistently match or mismatch predicates that the
clients used in their natural language. The no-match group
counseled normally. In comparing the matched and mismatched
predicates, the investigators found no significant differences in
how counselors were perceived and no significant differences in how
the total counseling relationship was perceived. An unexpected
finding, however, was that counselors in the no-match condition were
rated consistently higher than those in the other two experimental
In a second study, Dowd and Pety (1982), asked participants
to rate two recorded counseling scripts in which the counselor
consistently matched or mismatched the client's PRS predicates. The
authors found that the results revealed no predicate matching
effects but did reveal a significant effect on post-interview
willingness to see the counselor, based on sex of the counselor,
with male counselors rated higher.

Yapko (1981), exposed 30 participants to three taped hypnotic
inductions that differed only in predicate usage. Participants
obtained a greater degree of relaxation as measured by electromyo-
graph (EMG) when inductions matched their PRS than when they were
Falzett (1981), noted that the predicate usage method of
identifying PRS was not conducive to research purposes (assessment
of use of predicates revealed that only 3 out of 26 participants
were not predominantly kinesthetic). Therefore he recommended
employing the eye-movement technique, not predicates, for identify-
ing PRS. As a result, Falzett (1981), developed a research design
similar to that used by Dorn (1980), Dowd and Hingst (1981), and
Dowd and Pety (1982), in which clients were paired with counselors
who used similar and dissimilar PRS predicates in their natural
language. The results of this study indicated that clients perceive
counselors as significantly more trustworthy when the counselors use
PRS predicates that are similar to the clients. The PRS of the
clients had been determined by observing eye-movements and not by
predicates observed or heard in the clients' speech.
Two studies, (Gumm, Walker, and Day, 1982; Owens, 1977),
examined NLP methods of assessing PRS using predicates and eye-move-
ments inter-relatedly. Owens found agreement between eye-movements
and subject verbalization but not among any self-reports of
preferred sensory modes. Gumm et al., (1982), found no correlations
between predicates, eye-movements, and self reports. Gumm et
al., examined the PRS modalities of 50 right-handed college students

by assessing eye-movements, predicate usage in verbal reports, and
self reports of preferred modes. Results failed to reveal signifi-
cant agreement of the three assessment methods. In addition, each
assessment method was shown to be biased toward revealing a particu-
lar modality; the verbal report method (predicates) and the self
report method were biased toward revealing a kinesthetic modality;
whereas the eye-movement method indicated that the largest number of
subjects had auditory PRS's.
In an examination of literature regarding the assessment of
PRS, Dorn et al., (1983), noted the potential contribution that a
reliable method of identifying PRS would make to the literature on
NLP. According to Dorn et al.:
Existing research on NLP and the identification of PRS so
far has been disappointing. There does not seem to be a
reliable method of assessing the PRS; and this lack of a
reliable assessment method is the most disturbing factor
reappearing throughout NLP research, (p. 167)
Dorn et al., state that it is difficult to overestimate the
importance of this lack. If various NLP therapeutic strategies
(such as matching predicates to enhance rapport) are to be effec-
tive, then the PRS mode of the individual must first be reliably
Dorn et al., further state that existing research has been
valuable in that it has provided directions to pursue and method-
ological concerns that must be addressed:
Until now the major thrust of the proponents of NLP has
been training rather than research. Therefore, it would be
beneficial if more emphasis were directed to the latter
before NLP becomes just another approach to gain shelter
under the umbrella of therapeutic modalities, (p. 167)

In summary of this section, the inconclusive findings from
published studies regarding the predicate usage hypothesis speak for
themselves. The need for more accurate methods of assessing PRS
before using NLP therapeutically is obvious. Therefore a content
analysis method of PRS assessment would seem an appropriate response
to Dorn et al.'s (1983) methodological concerns.
Unpublished Research on Predicate Usage
(PhD. Dissertations)
Research findings on predicate usage described in disserta-
tions are as inconsistent and inconclusive as the findings described
in published studies. Of the studies described in dissertations,
some support the predicate usage hypothesis and the use of predicate
matching techniques, while others negate it or generate neutral
Brockman (1980), found that matching of clients' predicates
during the counseling process facilitated the clients' perception of
the empathic component of rapport. Brockman divided 24 subjects
into two groups; counselors matched predicates for one group and
mismatched for the other; then perceptions of empathy were measured.
Frieden (1981), found that predicate matching in psycho-
therapy produces increased eye contact but paradoxically increased
head-tohead distance. Frieden conducted a two subject case study.
Green (1979), found that trust between subjects and a trained
experimenter who used matching predicates were positively related.
Green divided 63 subjects into two groups: a group that worked with
a trained experimenter and a control group; then levels of trust

were measured.
Paxton (1980), found in counseling relationships that counse-
lors who use predicates reflecting a specific PRS will influence
more positive client perception of the relationship. Paxton taped
language samples of 48 clients and then analyzed and matched the
PRS1s identified.
Schmedlen (1981), found in interpersonal attractiveness that
PRS matching only makes a difference in perceived attractiveness,
when interacting with the opposite sex. In Appel's study, 143 sub-
jects rated the attractiveness of three male and three female taped
presenters who varied in their PRS presentation modalities.
Kraft (1982), in considering the therapy outcome of relaxa-
tion, found no main effects of matching PRS. Kraft classified 36
subjects, 18 males and 18 females, according to their PRS. Then the
subjects were exposed to three different PRS-oriented relaxation
tapes. Level of relaxation then was measured.
Rebstock (1980), found that training in PRS matching tech-
niques had no effect on the development of rapport between client
and counselor during initial counseling interviews. Six male and
six female counselors were trained in the use of PRS matching
techniques. A second group of six male and six female were not
trained. Then the rapport skills of the 24 counselors were
Cole-Hitchcock (1980), found no substantiation for the gen-
eralization that each person has a dominant PRS that can be identi-
fied. This study examined 33 subjects, transcribed interviews, and

videotaped eye-movements, and found no correlation between the two.
In summary of this section, as in published studies, there is
inconsistency regarding the efficacy of predicate-matching techni-
ques. These inconsistencies and contradictions regarding predicate-
matching suggest a need for refinement and replication of some of
these works. The Cole-Hitchcock (1980), dissertation perhaps is
worth replication, in that it questions the process of assessing and
identifying PRS accurately.
As the literature examination of Dorn et al., (1983) pointed
out, PRS assessment methodology should be a first concern of
academic research on NLP. Contrarily, most of the dissertation
studies were outcome oriented and focused on matching predicates in
a variety of contexts, rather than on evaluating PRS assessment
methodologies. Once again the need for improved content analysis
methodology is evident.
C. Discussion of Relevance of Literature
Reviewed to the Research Problem
The first section of this review pointed to communication
concerns regarding scientists and humanists and linked those
concerns to differences in cognitive style. The second section
linked cognitive style differences to subcultural differences in
language usage. The third section explicated a theoretical vehicle
for investigating language usage differences between cultural
subgroups, the predicate usage hypothesis. Based on the writings
reviewed in these three sections, the following hypothetical
assumptions are made:

1. Communication barriers between scientists and humanists
2. The communication barriers relate, at least in part, to
language usage differences.
BUT . .
3. The vehicle chosen to investigate these differences, the
predicate usage hypothesis, may be tenuous in terms of
efficacy and in need of further research.

A. Overview of the Chapter
The methodology for this project will be explained categori-
cally, in terms of:
1. Project Variables and Purpose of the Study;
2. Subject Group Design, that is selection and qualifica-
tion of the subjects;
3. Interview and Field Procedure; that is the formulation
of interview questions and conduct of the interviews;
4. Training of the Judges;
5. Content and Statistical Analysis of the Transcribed
6. Comparative Analysis of the Data; and
7. Limitations and Criticism of the Project Methodology.
B. Project Variables
Dependent Academic Orientation subgrouped as hard or
Variables: soft major.
Independent Predicate usage pattern subgrouped as
Variables: visual, auditory, or kinesthetic.
Graduating senior, right-handed, English as
native language.

The purpose of this study is to predict academic orientation
(the dependent variable) based upon the language pattern of the
subject (the independent variable). The project intends to explore
a correlation between the way a subject uses language and his/her
professional/academic orientation.
C, Subject Group Design
Participants for the study were selected randomly from the
student population at University of Colorado, Colorado Springs.
Before being accepted as a subject, each student was qualified based
upon the following criteria: (Appendix 1)
Senior Status. This qualification was based upon the assump-
tion that students' language patterns and their academic majors
might change through time, and that their academic majors might be
more established at the senior level and less established at lower
academic levels. Between the freshman and senior levels, students
might change their academic majors and/or begin to use language more
similar to others in their academic disciplines.
English as Native Language. This qualification assumed that
an individual who was not a native English speaker might use English
somewhat differently than native speakers and thus would confound
the study.
Right-handedness. This qualification was based upon other
published studies on NLP that acknowledged relationship between
right-handedness and cerebral processing and cerebral organization.

Supposedly left-handed people are cerebarally organized the opposite
of right-handed people. Including any left-handed subjects, there-
fore would have been a confounding variable for this study.
Naivete Regarding NLP and Willingness to Participate in a Ten
Minute Interview. Any prior knowledge of NLP and predicate usage
patterns would have prohibited the subject using natural language
patterns during the interview.
Time constraints also were a factor that might have influ-
enced the language used.
Declared Academic Major. This qualification categorized
students as subjects based on the academic focus of their major area
of study. Accepted subjects were categorized according to their
academic orientation as either hard or soft, i.e., scientists or
humanists. Any student whose academic orientation was not clearly
hard or soft was disqualified as a subject.
The design for this hard/soft majors categorization system
was based on similar subject group designs in three previous
studies, (Hilgard, 1969; Baken, 1969; Katz and Salt, 1981).
(Appendix 2) These three published works were the only studies
using a similar subject group design that could be located in
scholarly literature. However, the methods of designing the subject
groups in these three studies were problematic in several ways.
First, the studies drew many of their subjects from introduc-
tory or entry level classes. This allowed for the possibility that
the subjects might switch academic majors prior to reaching senior
status. Also, language patterns of seniors might be more firmly

established than freshmen or other lower academic levels.
Second, in terms of confounding variables contaminating the
subject groups, these three studies appeared to intuitively change
and arbitrarily define what was a hard or soft major. (Appendix 2)
Areas of study which were clearly scientific or clearly humanistic
presented no categorization problems; but some academic subjects,
particularly those of a social science nature, did present categori-
zation difficulties. Political science or communication, for
example, were categorized differently from one study to another.
Observing such a difficulty in categorizing social sciences,
this project chose to eliminate any possible confounding academic
majors in order to avoid contamination of either subgroup. Thus,
the following subject group categorization system for this project
was generated:
This categorization system is in accord with and wholly con-
sistent with all prior research undertaken. But note that this
system goes further and eliminates any possible confounding majors
such as nursing, political science, psychology, sociology, or com-
munication; thus the purely scientific and the purely humanistic are
isolated. The study of music and art as soft majors also are
eliminated, because of their possible relationship to the NLP
Engineering (any type)
Foreign Language

auditory or visual sensory modality. Although these constraints do
limit the number of potential subjects available, benefits to the
study are great in terms of eliminating contamination of the subject
The size of the subject group was dictated by the statistical
methodology used. The original intent was to obtain a minimum of 32
subjects, evenly divided between hard and soft majors. This number
is deemed adequate to establish statistical significance using the
analytical tools to be described herein. The final subject group
totaled 35 usable interviews which is adequate for the required ten
respondents per independent variable (visual, auditory, kines-
thetic) .
Subjects were randomly approached for participation in this
project in the Student Center at the University of Colorado,
Colorado Springs. Each subject first was individually qualified as
a senior, right-handed, native English speaker, and naive regarding
NLP, based on their own self-reports of these factors. Subjects
were only accepted for participation if they were considered hard or
soft majors, according to the operational definition of the sub-
groups by the categorization system. No data or details other than
the qualification factors were discussed that could have biased the
subjects' later responses to interview questions. All subjects who
met the qualifications were advised of a time and location for their
individual interviews.

Six interview questions had been formulated and pre-recorded
on audiotape. These six questions were recorded by both a male and
female voice, randomly alternating which questions were asked by the
female voice and which by the male voice. Four different sets of
the same six questions were recorded with both voices carefully
controlled to eliminate variation in tempo, tonality, and inflec-
tion, which might have biased the subjects' responses. The ques-
tions were worded simply to eliminate any leading language which
also could have biased responses. (Appendix 1) The six questions
focused on a past, present, and future event, and on a visual,
auditory, and kinesthetic experience. The other two sensory
modalities, i.e., olfactory and gustatory, were not included, as
these modalities are not considered predominant in the language
patterns of our culture.
Prior to commencing each interview, a brief warm-up discus-
sion period was conducted to encourage the subject to relax and to
use his/her most natural language patterns. The topic of discussion
for the warm-up period was the operation of the recording for the
interview. The subjects were instructed regarding the use of the
equipment, and then were encouraged to be relaxed in their responses
and to avoid simple yes/no responses. The interview questions were
on one tape machine, controlled at will by the subject, and his/her
responses were recorded on a second tape machine. Each subject was
assigned a code number that was recorded at the beginning and end of
the interview on the tape. The same code number was placed on the
subject's interview questionnaire.

During the interview, the interviewer remained present but
did not comment regarding the subject's responses. The subject
controlled the tape equipment, answering all six questions at
his/her own pace. The interview was audiotaped to be transcribed
later for content analysis.
Following each interview, the subjects filled out a question-
naire related to personal, demographic, familial, personality, and
academic aptitude factors. (Appendix 3) The purpose of this ques-
tionnaire was to supply additional data on each subject to examine
later for significant findings or correlations.
After filling out the questionnaire, each subject was
thanked, debriefed, and supplied with a verbal description of the
general nature of the research project. If the subject had interest
in receiving results of the study, that interest was noted on the
E. Training of the Judges
In order to reduce sex-related content bias, a male and a
female judge, naive regarding NLP and PPUP, were trained to use the
content analysis system designed for this project. Based on two NLP
predicate lists, (Appendix 4), the content analysis judges were
trained to identify predicates (verbs, adverbs, adjectives) as
either visual, auditory, or kinesthetic in connotation (or not-
categorizable). The not-categorizable or dump category was intended
for any predicates that had neither visual, nor auditory, nor
kinesthetic connotation. The judges were given no information

regarding expected theoretical differences in predicate usage, while
being trained to identify predicates. After this initial training,
each judge was given a copy of a transcribed interview in which all
the predicates were underlined. Both judges separately identified
the predicates as either visual, auditory, or kinesthetic, or not-
categorizable. The judges' analyses of this first interview were
compared. They attained 84% agreement regarding the sensory nature
of the underlined predicates. The judges then did two more trans-
cribed interviews in the same manner and attained approximately the
same level of agreement or inter-rater reliability. Both judges
then met with a certified NLP practitioner to ask questions and
discuss some areas of dispute and to ascertain more accurately what
NLP deems to be a visual, auditory, or kinesthetic predicate.
Following that meeting, the-two judges analyzed one more transcribed
interview and attained a level of 90% agreement regarding the
classification of predicates. This level of inter-rater reliability
is adequate according to the content analysis reliability standards
of Holsti (1969). Therefore the two judges were deemed ready to
analyze actual interviews for the study.
F. Content Analysis of Transcriptions
There was no content analysis model available to replicate in
published studies. All previous published studies utilized an
intuitive hearing analysis process for identifying either a visual,
auditory, or kinesthetic subject, rather than a content analysis

The purpose of using content analysis for this project was to
examine the transcribed interviews of the subjects in terms of pre-
dominant predicate usage patterns, that is, in terms of the
quantitative frequency of visual, auditory, and kinesthetic words.
First, the transcribed interviews were analyzed for predicates. All
the complete predicates (all verbs, adverbs, and adjectives or
verbial, adverbial, and adjectival phrases) were identified and
underlined by a research assistant trained to identify these parts
of speech. Then each trained judge, the male and female, were given
copies of each transcribed interview in which the predicates were
underlined. Both judges separately identified all of the predicates
in each transcription as V_, A, K, or not-categorizable. The
tallying of predicates by each judge was recorded on a judge's
report sheet. (Appendix 5) The total number of each type of
predicate was tabulated for each interview along with the total
number of predicates altogether for that interview. Thus each
subject's use of visual, auditory, and kinesthetic predicates could
be calculated in two ways; that is by actual frequency and a
percentile frequency (i.e., the precentage of the total predicates
that indicated a particular modality.)
G. Statistical Analysis of Data
The numerical data generated by content analysis of the sub-
jects' transcriptions were evaluated for each subject utilizing the
Pearson Correlation Coefficient Process. Means and standard devia-
tions for each judge for each predicate category (V, A, K) were

generated. Then correlations were drawn between each judge's predi-
cate identification percentages (V, A, K) and academic orientation
(hard vs. soft), using Pearson Correlation Coefficients. The level
of agreement between the two judges, their inter-rater reliability
in terms of predicate identification, was calculated as a final
analytical step.
H. Comparative Analysis of Data
The audiotapes of the subjects' interviews were given to two
trained NLP practitioners, a male and female, who listened to the
interviews utilizing an intuitive hearing process to identify the
number of visual, auditory, or kinesthetic predicates for each sub-
ject. The purpose of this comparative analysis was to contrast the
intuitive identification process of the NLP trained practitioners to
the results that had been generated by the content analysis process
of the trained judges. The statistical analysis process that was
applied to the data of the content analysis judges also was applied
to the data generated by the NLP judges. Then the two sets of
judges were treated as separate panels and were compared for their
inter-reliability and inter-panel reliability.
I. Limitations and Criticism of the Project Methodology
The methodology is sound of design in terms of its ability to
ascertain PPUP of graduating seniors. However, facets of the
methodology raised questions for redesign and alternative approaches
in further studies:

a. Formulation of Subject Group
Although using graduating seniors rather than entry level
students was advisable, generalizing from this subject group to
scientists and humanists in the real world is questionable. A
refined study might draw subjects from the real world, scientists
and humanists who have been working actively in a professional
capacity. A recommended control variable also might be employment
as a scientist or humanist, that is, in a hard or soft profession,
for a minimum of five years.
b. The Interview Process
When the subjects spoke to a tape recorder rather than to a
live interviewer, their responses seemed abbreviated and not as
natural as would be preferred for this study. A refined study might
use two live interviewers, a male and a female, trained to control
their vocalics for variation in tempo, tonality, and inflection.
Rather than just ten minutes, a refined study might conduct a longer
interview, of a half-hour to an hour, in order for subjects to relax
and thus use more natural language patterns. Then a ten-minute
segment of the interview could be extrapolated for purposes of con-
tent analysis. That ten-minute segment might exhibit more relaxed
and natural language usage by the subject.
Other than these two areas of criticism, the project
methodology proves relevant and purposeful for accomplishing the
study's goal.

A. Introduction to the Findings
Considerable data and findings were generated by the method-
ology described. Some of the findings related directly to the pro-
ject's original hypothesis of a correlation between predicate usage
and academic orientation; while other findings, not originally
intended to be explored, relate to the efficacy of the predicate
usage hypothesis itself. Although the earlier stated purpose of
this project was not to explore the predicate usage hypothesis,
initial test results dictated some adaptation in research focus.
Therefore some inquiry into the predicate usage hypothesis was made,
by adding the comparative analysis of the data by NLP-trained
judges, in addition to content analysis judges. The results of both
panels of judges are explicated in three tables (1,2,3) presented
herein. In those tables, the results of the four judges are
presented as: Panel I (the two judges, #1 and #2, who performed
content analysis of predicates on transcribed interviews); and Panel
II (the two judges, #3 and #4, who used the NLP intuitive listening
process of identifying predicates on audiotapes of the interviews).
The tables of findings (1,2,3) indicate the following

B. Summary of the Findings
Total Predicates Predicates by Sensory Modality
Mean S.D. Visual Mean S.D. Auditory Mean S.D. Kinesthetic Mean S.D.
Panel I (Content Analysis Judges)
Judge #1 152.3 81.51 16.5 12.64 11.2 8.38 68.7 32.54
Judge #2 155.7 83.09 19.5 11.79 11.9 11.36 60.1 41.74
Panel II (NLP Judges)
Judge #3 153.6 82.00 5.9 3.98 7.7 5.00 6.8 3.84
Judge #4 153.6 82.00 3.6 3.20 4.7 3.05 9.0 4.34
Table one is an analysis of the four judges in terms . of the mean
frequency of visual, auditory, or kinesthetic predicates each judge
identified in the subjects' transcribed or audio-taped interviews.
Three out of the four judges identified more kinesthetic predicates
than visual or auditory.

(Utilizing Pearson Correlation Coefficients, Positive or
Negative Degree of Correlation Evaluated at a .05 Level of
Predicates by Sensory Modality
Visual/ Auditory Kinesthetic/
Hard Major Soft Major
Individual Judges:
Content Analysis Judges
1 -0.2489 -0.1225 +0.2671
(p<.075) (p<.242) (P<.060)
2 -0.0625 -0.2707 +0.3366
(p< .361) (p<.058) (p<.024)*
NLP Judges
3 -0.3302 +0.0082 +0.3425
(p<.046)* (p<.484) (p<.040)*
4 +0.0887 -0.3167 +0.2014
(p<.320) (p<.044)* (p<.143)
Panel I:
Judges 1 & 2 -0.2039 -0.2156 +0.3435
(p<.120) (P<.107) (p<.022)*
Panel II:
Judges 3 & 4 -0.2335 -0.1134 +0.3154
(p<.121) (p<.287) (p<.055)*
All Judges:
1,2 ,3,& 4 -0.2465 -0.1436 +0.3747
(p<.108) C P <-237) C p<.027)*

TABLE 2 (continued)
Table two is an analysis of the four judges' scores on
predicate identification correlated with academic orientation of
subjects. Three correlations are presented:
1. The scores for each of the four judges are correlated
individually with academic orientation.
2. The scores of judges 1 and 2 are combined and correlated
with academic orientation (as Panel I); and the scores of
judges 3 and 4 are combined and correlated with academic
orientation (as Panel II).
3. The scores of all four judges (1,2,3,4) are combined and
correlated with academic orientation.
Out of the three correlative analyses, the only correlations
that are positive and statistically significant (p<.05) are those
correlating kinesthetic predicates with academic orientation (Hard
vs. Soft Majors). Specifically, kinesthetic predicates indicate
positive correlation with soft majors in all three analyses. This
finding is expected considering the results in Table 1, that is the
greater use of kinesthetic predicates by the subjects and the ten-
dency of three of the four judges to identity more kinesthetic
predicates. The only positive correlation of visual predicates, by
Judge 3, was not statistically significant.
The only negative correlations that were statistically
significant were in one predicate category for Judge 3 and one
predicate category for Judge 4.

(Part A) OF PANEL I (Judges 1 and 2) and PANEL II (3 and 4)
(Utilizing Pearson Correlation Coefficients Negative Degree of Correlation Evaluated at Significance) , Positive or a .05 Level of
Panel I: (Content Analysis Judge)
J2/Visual J2/Auditory J2/Kinesthetic
Jl/Visual +0.1784 (p<.153) +0.0417 (p<.406) -0.658 (p<.354)
Jl/Auditory -0.929 (p<.298) +0.8217 (p<.000)* -0.1547 (p<.187)
Jl/Kinesthetic +0.3124 (p<.034)* -0.2771 (p<.054) +0.5301 (p<.001)*
Panel II: (NLP Judges) J4/Visual J4/Auditory J4/Kinesthetic
J3/Visual +0.4356 (p<.012)* -0.2687 (p<.088) -0.0973 (p<.315)
J3/Auditory -0.1322 (p<.255) +0.4969 (p<.004)* -0.3205 (p<.052)
J3/Kinesthetic -0.3355 (p<.044)*. -0.1915 (p<.169) +0.4107 (p<.017)*
(Utilizing Fisher r to z Transformation of the Correla-
tion Coefficients from Part One)
Panel I (Content Analysis Judges) J1&J2, Visual = 0.1799 Panel II (NLP Judges) J3&4, Visual = 0.4673
J1&J2, Auditory = 1.1630 J3&4, Auditory - 0.5453
J1&J2, Kinesthetic = 0.5901 J3&4, Kinesthetic = 0.4356
1.9330/3 = 0.6443 = 0.5676 Transformed (z Score) (r Score) 1.4482/3 = 0.4827 = 0.4478 Transformed (z Score) (r Score)

TABLE 3 (continued)
Table three is an analysis of the inter-rater reliability
levels of the two types of judges, the content analysis judges
(Panel I) and the NLP-trained judges (Panel II).
In Part A, correlations are made within each panel of two
judges in terms of their tendency to similarly identify predicates
(V,A,K). Both panels of judges demonstrated positive correlation
regarding identification in the three categories of predicates.
These positive correlations are statistically significant (p<.05),
with the exception of the Panel I visual category. The only
negative correlation that is significant (p<.05) is for Panel II
correlating visual and kinesthetic predicates for Judges 3 and 4.
Comparing the positive correlations of the two panels for
each category, Panel 11s correlations are higher for auditory ( + .82
vs. +.50) and for kinesthetic ( + .53 vs. .41); Panel 111 s correla-
tion is higher for the visual category ( + .44 vs. +.18). These
results indicate that the content analysis judges generated a higher
level of inter-rater reliability (in two categories of predicates)
than the NLP-trained judges (in one category).
In Part B, this higher level of inter-rater reliability for
Panel I was verified by calculating and comparing average correla-
tions for each panel utilizing Fisher's r to z transformation of the
correlation coefficients from Part A. Panel I's score (.57) was
higher than Panel II's (.45).
The findings in the three tables fail to support the present
project's original hypothesis regarding predicate usage and academic
orientation. Further, the findings call into question the predicate
usage hypothesis itself. Discussion of these two major project
results is in the next and concluding chapter.

An initial explanation is in order regarding the project
findings described. It is obvious that the findings speak to two
research questions. The original project hypothesis regarding cor-
relation of predicate usage patterns and academic orientation is
considered. But the efficacy of the predicate usage hypothesis is
addressed also. This investigative extension of the original
project hypothesis was precipitated by the findings of the original
field investigation and content analysis process. When the first
two judges completed their content analysis task of the subjects'
transcriptions, the subjects' bias toward kinesthetic predicate
usage and a kinesthetic modality was obvious. That bias toward one
modality called into question the validity of predicate usage as a
means of identifying a subject's preferred sensory modality. And if
the subjects demonstrated little variation regarding the three
predicate categories (V,A,K), then correlation with academic major
became a moot point.
Inquiry into academic literature, (Falzett, 1981; Gumm et
al., 1977), provided some support for questioning the predicate
usage hypothesis. Both Falzett and Gumm et al.,, had concluded that
predicates were unreliable as predicators of subjects' primary

representational systems. As a result, the original project was
extended to investigate the predicate usage hypothesis itself. The
question of whether predicates could or could not be identified
accurately became a research question of relevance to this study.
Therefore audiotapes of the subjects' interviews were provid-
ed to two NLP practitioners, a male and a female, for purposes of
predicate identification. Then both sets of data, from the content
analysis judges and the NLP judges, were analyzed statistically in a
similar manner. The results of those analyses are presented in the
tables herein and are now discussed.
A. Discussion of the Test Results
The first set of test results to be discussed relate to the
four judges' mean scores on predicate identification by sensory
modality category (V,A,K), (Table 1). Judge 1 identified more than
four times as many kinesthetic predicates by comparison to visual or
auditory. Judge 2 identified three times as many kinesthetic as
visual or auditory. Judge 4 identified two to three times as many
kinesthetic. Only Judge 3 did not indicate a bias toward kines-
thetic predicate identification; and her scores for all three
categories were so close (5.9, 7.7, 6.8) that her bias toward an
auditory modality is not significant. As stated, these results are
in accord with the cited studies of Falzett (1981), and Gumm et
al., (1977). Both Falzett and Gumm' et al., found subjects use of
predicates so biased toward a kinesthetic modality, that predicate
identification was deemed not a viable method of assessing subjects'
PRS. These results have obvious ramifications for this project and

its hypothesis, and for other NLP outcome studies.
Regarding this project, these findings support a null
hypothesis of no correlation between predicate usage and academic
orientation. If most subjects are highly kinesthetic in predicate
usage, then it is not expected that any subjects will be highly
visual and correlate positively with the hard major academic
orientation. Regarding other published NLP outcome studies, the
results of those studies become, as a result of these findings,
somewhat suspect. As described in the literature review, many NLP
studies have measured outcomes such as trust, enhanced rapport,
etc., as effected by matching subjects' predicates. The matter of
first accurately identifying or assessing the subjects' favored
predicate usage patterns is not a primary question. This short-
coming in NLP research, the lack of a reliable method for assessing
PRS, was outlined in the cited writings of Dorn et al., (1983).
Based on Dorn's writings, and the kinesthetic bias indicated in
these finding, NLP outcome studies and their results are considered
The second set of test results to be considered are those
related to the project's original hypothesis of a correlation
between visualpredicates and hard majors and kinesthetic predicates
and soft majors (Table 2). There was no significant positive
correlation of visual predicates with academic orientation (p<.05)
by any of the judges, either individually, collectively as panels,
or collectively as a group. There was positive correlation of
kinesthetic predicates with academic orientation (p<.05) in all

three analytical correlations, individually by judges, as panels,
and collectively as a group. The only exception to positive
significant correlation of kinesthetic predicates was Judge 4,
individually, who indicated positive correlation but it was not
significant. The only negative correlations that were significant
were for Judge 3 on visual predicates and Judge 4 on auditory
These results were to be expected considering the subjects'
bias toward kinesthetic predicate usage. There is significant posi-
tive correlation between kinesthetic predicates and soft majors, but
no positive correlation between visual predicates and hard majors..
These results do not indicate that language usage differences
between hard and soft majors, (ise., between scientists and human-
istics), should not be investigated, however. Rather, they suggest
that the methodology chosen to investigate those differences proved
lacking; that is, the predicate usage hypothesis as the chosen
methodological investigative tool, proved problematic.
A final set of test results to be discussed relate to the
inter-rater reliability level of content analysis judges as opposed
to NLP-trained judges. This inquiry and these tests were added due
to what was learned about the methodological inadequacies of predi-
cate usage identification and assessment. Surprisingly, both pro-
cesses (content analysis and the NLP method) indicated positive
correlation between judges regarding inter-rater reliability on the
three predicate categories (V,A, and K). The inter-rater relia-
bility level of both sets of judges was significant (p<.05) for most

correlations. The only exception was the content analysis process
(Panel I) for the visual category. Average correlations of the
inter-rater reliability levels of the two panels of judges indicated
that Panel I (content analysis) had only a slightly higher level of
reliability than Panel II (NLP-trained). These results speak well
for NLP training in that the NLP judges (Panel II) tended to
identify predicates similarly. But a question is raised by these
results if juxtaposed against other findings from other studies,
(Falzett, 1981; Gumm et al., 1977). If language patterns are
significantly biased toward kinesthetic predicates, then the NLP
training which focuses on identifying three different predicate
usage patterns (V,A,K) is questionable. And the results of these
inter-rater reliability tests indicate that these two NLP judges are
trained to identify for all three patterns, which may not be a valid
construct. As a part of NLP training, NLP practitioners learn to
identify subjects as visual, auditory, or kinesthetic based on
predominant use of one particular predicate type. The results of
this study question that training.
SUMMARILY, when reviewing all of the test results described,
two over-riding conclusions may be drawn:
1. that the original project hypothesis of a correlation
between predicate usage and academic orientation is not
supported; and
2. that the grounding methodology chosen to investigate the
project hypothesis is called into question.
Rather than viewing these two conclusions negatively, further
academic inquiry is suggested.

B. Recommendations for Further Study
In addition to replication, two other recommendations for
further academic investigation are suggested by the test findings of
this project:
First, it is obvious that a primary tenet of NLP theory, the
predicate usage hypothesis, has shortcomings and is in need of fur-
ther research. It also is obvious that the focus of further
research should not be on outcome or effects studies, that use
predicatematching techniques. Rather, methods for assessing
predicates, and the efficacy of the predicate usage hypothesis,
should be investigated. As noted in the literature review, there is
an abundance of published and unpublished research involving the
predicate usage tenet. But most of these studies assume as a given
that predicates can be matched and that there will be some variation
in predicate usage (V,A, or K). Based on that assumption, predi-
cates are matched in a variety of contexts and the effects of the
matching are measured. Only two published studies, (Falzett, 1981;
Gumm et al., 1977) questioned the predicate identification process
and both of those studies determined that predicate usage is biased
toward a kinesthetic modality. Replication of either of these
studies, particularly Gumm et al., is recommended.
Considering the primary role that predicate usage occupies in
the NLP theoretical model, any further testing of predicate iden-
tification is recommended. Such testing might be modeled after this
project's methodology, but use considerably more judges. Two panels
of judges could analyze subject's language patterns. One set of

judges would perform content analysis on transcriptions of audio-
tapes and a second set of NLP-trained judges would react to listen-
ing to the audiotapes. If such testing revealed significant bias
toward a kinesthetic modality by either or both sets of judges, then
outcome studies using predicate-matching techniques would be
invalidated. Further, if a grounding tenet of the NLP model (the
predicate usage hypothesis) is found questionable, then the entire
model also is somewhat suspect.
A second recommendation for further academic inquiry, as a
result of this project, relates to the original project hypothesis.
That hypothesis was based upon the philosophical concerns of
C.P. Snow (1963), regarding a communication gulf between scientists
and humanists. Although this study did not find differences in
predicate usage between scientists and humanists, the communication
gulf of concern to C. P. Snow still exists, and should be researched
and tested. Even though the methodological model chosen by this
project proved problematic, there is no indication that other langu-
age and communication style differences between scientists and
humanists are not a concern. The nature of those differences is yet
to be discovered.
Therefore this project points to further investigation of
communication style differences between scientists and humanists,
perhaps without the use of rigid models such as NLP and the predi-
cate usage hypothesis. The use of such models defines what is to be
observed and possibly screens out potentially interesting factors
that might be of interest to the research observer. It might prove

worthwhile, simply to observe the verbal and nonverbal communication
styles and patterns of scientists and humanists, rather than be
confined by one's own limiting hypothesis. Then after observing any
language or communication style differences, of a verbal or non-
verbal nature, the observed differences could be investigated
further. Such an investigative approach might prove the best
methodology for addressing. .
. . the polarization that is sheer loss to us all,
practical, intellectual, and creative loss. (p. 11)
C. P. Snow

Asuncion-Lande, Nobleza. "Language Theory and Linguistic
Principles." International and Intercultural Communication
Annual, VII, 1983, 253-257.
Bandler, R. & Grinder, J. Structure of Magic, I. Palo Alto,
Calif.: Science & Behavior Books, 1975.
Bandler, R. & Grinder, J. Frogs Into Princes. Moab, Utah: Real
People Press, 1979.
Bernstein, Basil. Class, Codes, and Control. Vols. I, II, & III.
London: Routledge-Kegan Panel, 1971.
Bogen, Joseph E. "Some Educational Implications of Hemispheric
Specialization." The Human Brain, Englewood Cliffs, New Jersey:
Prentice-Hall, 1977.
Cameron-Bandler, L. Solutions. San Rafael, Calif.: Future Pace,
Inc., 1985.
Dilts, Robert. Neurolinguistic Programming. Santa Cruz, Calif.:
Behavioral Technology, 1980.
Dilts, R., Grinder, J., Bandler, R., Bandler, L.C., & Delozier, J.
Neurolinguistic Programming I. Cupertino, Calif.: Meta Publica-
tions, 1980.
Dorn, F.J. The Effect of Counselor Client Sense Modality Similarity
on Counselor Attractiveness. Unpublished doctoral dissertation,
Iowa State University, 1980.
Dorn, F.J., Brunson, B.I., and Atwater, M. "Assessment of Primary
Representation of Systems with Neurolinguistic Programming:
Examinational Preliminary Literature." American Mental Health
Counselors Association Journal, Oct. 1983, 161-168.
Dowd, E.T. & Hingst, A. Counselor Predicate Matching: An in Vivo
Demonstration of Effectiveness. Paper presented at the American
Psychological Association Annual Convention, Los Angeles, Calif.,
Dowd, E.T. & Pety, J. "Effect of Counselor Predicate Matching on
Perceived Social Influence and Client Satisfaction." Journal of
Counseling Psychology, 1981, 28, 305-308.

Falzett, W. "Matched Versus Unmatched Primary Representational
Systems and Their Relationship to Perceived Trustworthiness in a
Counseling Analogue." Journal of Counseling Psychology, 1981,
28, 305-308.
Folb, Edith A. "Vernacular Vocabulary: A View of Interracial
Perceptions & Experiences." Intercultural Communication: A
Reader, Samovar & Porter, Belmont, Calif.: Wadsworth Publishing
Co., 1982.
Gonzales, R.R. & Roll, S. "Relationship Between Acculturation,
Cognitive Style & Intelligence." Journal of Cross-Cultural
Psychology, Vol. 16, No. 2, .June 1985, 190-205.
Gordon, David. Therapeutic Metaphors: Helping Others Through the
Looking Glass. Cupertino, Calif.: Meta Publications, 1978.
Grinder, J. & Bandler, R. Structure of Magic, II. Palo Alto,
Calif: Science & Behavior Books.
Gumm, W.B., Walker, M.K. & Day, H.D. "Neurolinguistics
Programming: Method or Myth." Journal of Counseling Psychology,
1982, 29, 327-330.
Harmon, R. & O'Neill, D. Neurolinguistic Programming for
Counselors. Personnel & Guidance Journal, 1981, 59, 449-453.
Hodges, John. Harbrace College Handbook, 7th Ed. New York:
Harcourt, Brace & Jovanovich, 1972.
Hoijer, Harry. "The Sapir-Whorf Hypothesis." Language in Culture.
Chicago: University of Chicago, 1954.
Holsti, O.R. Content Analysis for the Social Sciences and Humani-
ties. Reading, Mass.: Addison-Wesley, 1969.
Katz, J. & Salt, P. "Differences in Task & Use of Language, A Study
of Lateral Eye-movement." Perceptual & Motor Skills, 52, 1981,
Lankton, S. Pract i cal Magic. Cupertino, Calif.: Meta
Publications, 1980.
Owens, L.F. An Investigation of Eye Movements and Representational
Systems. Unpublished Doctoral Dissertation, Ball State
University, 1977.
Samovar, Larry & Fred Sanders. "Language Patterns of the
Prostitute: Insight into a Deviant Culture." Intercultural
Communication: A Reader, Samovar & Porter, Belmont, Calif.:
Wadsworth Publishing Co., 1982.

Sandell, Rolf. Linguistic Style and Persuasion. New York:
Academic Press, 1977.
Sapir, Edward. "Conceptual Categories in Primitive Languages."
Science, 1931.
Shaw, D.L. Recall as Effected by the Interaction of Presentation
Representational System and Primary Representational System.
Unpublished Doctoral Dissertation. Ball State University, 1977.
Shneiderman, Ben. Software Psychology. Cambridge: Winthrop
Publishers, Inc., 1980.
Snow, C.P. Two Cultures: A Second Look. Cambridge: University
Press, 1963.
Staley, C.M. "Sex-Related Differences in the Style of Children's
Language." Journal of Psycho-linguistic Research, 11, 2, 1982,
Thomason, T., Arbuckle, T. & Cady, D. "Test of Eye-movement Hypo-
thesis of Neurolinguistic Programming." Perceptual and Motor
Skills, 1980, 51, 230.
Whorf, Benjamin. Collected Papers on Metalinguistics. Washington,
D.C., Dept, of Defense, Foreign Service Institute, 1952.
Yapko, M. "The Effect of Matching Primary Representational System
Predicates on Hypnotic Relations." American Journal of Clinical
Hypnosis, 1981, 23,3.

Appel, R.R. Matching of Representational Systems and Interpersonal
Attraction. PhD Dissertation, United States International
University 1983.
Brockman, W.P. Empathy Revisited: The Effect of Representational
System Matching on Certain Counseling Process and Outcome Vari-
ables. PhD Dissertation, The College of William and Mary,
Virginia, 1980.
Cole-Hitchcock, S.T. A Determination of the Extent to Which a Pre-
dominant Representational System Can be Identified Through
Written and Verbal Communication and Eye Scanning Patterns. PhD
Dissertation, Baylor University, 1980.
Frieden, Fredrick P. Speaking the Client's Language: The Effects
of Neurolinguistic Programming (Predicate Matching) on Verbal and
Non-verbal Behaviors in Psychotherapy. A Single Case Design.
PhD Dissertation, Virginia Commonwealth.University, 1981.
Green, M.A. Trust as Effected by Representational System
Predicates. Ph.D. Dissertation, Ball State University, 1979.
Kraft, William A. The Effects of Primary'Representational System
Congruence on Relaxation in a Neuro-Linguistic Programming
Model. PhD Dissertation, Texas A & M University, 1982.
Paxton, L.D. Representational Systems and Client Perception of the
Counseling Relationship. PhD Dissertation, Indiana University,
Rebstock, Marc E. The Effects of Training in Matching Techniques on
the Development of Rapport between Client and Counselor During
Initial Counseling Interview. PhD Dissertation, University of
Missouri-Kansas City, 1980.
Schmedlen, George W. The Impact of Sensory Modality Matching on the
Establishment of Rapport in Psychotherapy. PhD Dissertaion, Kent
State University, 1981.

y pisxjv-i

now a senior?
your declared academic major?
sh your native language?
now anything about NLP? . . ,
u be willing to participate in a ten-minute interview and
1 out a brief questionnaire?
detail a recent vacation or trip.
detail your personal thoughts about being interviewed.
detail, what you think you will be doing five years from now.
detail your favorite movie,
detail a recent memorable conversation.
detail how you are when very excited.
S i_______________________________LOCATI ON s________
DATEs ____ _____
1 1


i i
i j i i > (

lgard, J. R., M.D., Personality and Hypnosis. 2nd Edition.
Lversity of Chicago Press, 1979, p. 157.
Group 1 Group 2 Group 3
Social Studies Biological Sciences Humanities
t Physical Sciences
k a n, P. "Hypnotizability, Laterality of Eye Movements, and
ictional Brain Asymmetry." Perceptual and Motor Skills. 28, (1969),
political science
international relations
tz and Salt. "Differences in Tas
teral Eye Movement." Perceptual
science-oriented fields
political science
: and Use of Language, A Study of
and Motor Skills. 52, (1981), 995-
music, education, religion
siness, Marketing, Accounting, Psychology, Sociology, Political
ience, Computer Science, Guidance and Counseling, Anthropology,
ology, Education, Music and Fine Arts, Communication and

^ sc
!rof ssion

ie Orientation or Nationality^
of ss si on or Career (Subject s)^
r favorite sports or leisure activist
activities and/or sports that actually Ciceupy most of your
ur average GPA in scientific type subjects? high_jpedt*w ^,
ur average GPA in humanistic type subjects? high .mad_low
arural environment ,> do you. prefer to look or do you prefer ts*
uch that which you perceive? eg,, reeks* leaves, barke water?
touch look_____look & touch
it environrmnt0 how impert.&a
our early family experiences and
1lowing factors?
Very Important-Somewhat XBmert&nt-Jlefc lraportant*=0'p
i t trac t a. van s s

'~s; m\ Paintings
g Emotion by
and Hugging
ppearanee Of s
t Family Auto
@ House
Lothes {, etc *
aa rij3aairavuKsrac,3isiaa

ook forward to
ire am
te as
Visual Memory
Watched Concise
Observed Describe
Recall Noticed
Picture Appears
Note See i
Fantasy Look
Dream Recollect
Recognize Observation
Remember Distinct Clear
Auditory Memory
Criticize .So to speak
Sounds . Sounds crazy
Said Listened
Talked Harmonious
Rings a Speak
bell Tone/n J f
Clicks Dissonata
Popped in Buzz
my head Vocalize
Related Tap a cord
Stated Consider
Explained Insult
Recalled Admit
'eeling) * Auditory Dialogue
l On top of ^ (Same as Auditory
way Rough Memory)
Smooth '~"-~
kome Get off my back
kind Cramp my style
Grinding on my mind
Lift a burden
Stick it in the mail
't walk
Skibble over
Get it together
1 dig

a ted
aeons tic
a loud
pea 1
thund er

as ive
s thetlc
pervad ing
s trong

f t Uf 1 iww MW a.
talNo, Visual Auditory Kinesthetic Dump PPUP HARD/SOFTACAD.MAJOI

\ . y
i |

. !
- i

1 1 i