Citation
Toward a rhetoric of graphics

Material Information

Title:
Toward a rhetoric of graphics a tagmemic analysis of levels of abstraction in technical illustrations
Creator:
Osman-Jouchoux, Rionda
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
viii, 142 pages : illustrations ;

Subjects

Subjects / Keywords:
Technical illustration ( lcsh )
Tagmemics ( lcsh )
Tagmemics ( fast )
Technical illustration ( fast )
Genre:
theses ( marcgt )
non-fiction ( marcgt )

Notes

General Note:
Department of English
Statement of Responsibility:
Rionda Osman-Jouchoux.

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:
221452443 ( OCLC )
ocn221452443

Full Text
TOWARD A RHETORIC OF GRAPHICS: A TAGMEMIC ANALYSIS
OF LEVELS OF ABSTRACTION IN TECHNICAL ILLUSTRATIONS
by
Rionda Osman-Jouchoux
B.A., Ohio State University, 1970
lie., University de Paris X, 1983
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 Science
Technical Communication
1991


This thesis for the
Master of Science
degree by
Rionda Osman-Jouchoux
has been approved for the
Program of
Technical Communication
by
Charles E. Beck
R. Scott Grabinger
Daniel Montano
Date


Osman-Jouchoux, Rionda (M.S., Technical Communication)
Toward a Rhetoric of Graphics: A Tagmemic Analysis of Levels of
Abstraction in Technical Illustrations
Thesis directed by Associate Professor Charles E. Beck
Within the framework of a tagmemic analysis, a context-
specific study of the abstraction of graphics provides insight into the
effective use of graphics in popular scientific literature. This analysis
allows us to describe the rhetorical context, the graphics themselves
as rhetorical elements, and the interaction of the graphics with the
text. In designing graphics, technical communicators face decisions
about the content of graphic material and the presentation of that
content. While considering graphics as rhetorical elements, we can
approach their study through a discussion of the abstraction of form
and content of the graphic relative to the rhetorical situation. Using a
working definition, we may describe abstraction, both of content and of
form, as continua extending from concrete to abstract. To describe the
rhetorical situation, classical stasis theory allows us to discern the
issues addressed and the organization of a text.
The form and content of this abstract are approved. I recommend its
publication.
Signed
Charles E. Beck


TABLE OF CONTENTS
Figures.............................................. iv
Tables.............................................. vii
Acknowledgements................................... viii
CHAPTER
1. INTRODUCTION......................................... 1
The Problem of Designing Effective Graphics........ 2
Seeking Solutions in Rhetorical Studies............ 5
Organization of this Thesis.......................... 6
2. THE QUESTION OF ABSTRACTION.......................... 8
Visual Elements...................................... 9
The Concrete-Abstract Continuum..................... 10
3. TAGMEMICS: A METHODICAL ANALYSIS.................... 16
Tagmemics: A Translators Tool...................... 17
Examples of Tagmemic Analyses....................... 20
Bar Graphs vs. Line Graphs........................ 20
Analysis of Table vs. Bar Graph.................. 24
Funding Table Conclusions.................... 33
Funding Bar Graph Conclusions................ 39
Summary
39


4. ANALYZING GRAPHICS FROM POPULAR SCIENTIFIC
LITERATURE........................................... 40
Spinning Comets by Fred L. Whipple................. 41
A Mechanical Analysis............................. 4L
Inventory of the Graphics......................... 45
An Expert Speaks to Lay People....................... 45
Brodies Drawing of Coggia's Comet................ 49
Schmidt's Drawing of Tebbutts Comet.............. 53
Struve's Drawing of Donatis Comet of 1858........ 57
Schmidt's Drawing of Donatis Comet............... 60
Photographs of P/SW1.............................. 64
Zdenek Sekanina................................... 68
Graph of Enckes Comet............................ 71
Jet Force Diagram................................. 75
Wobble Cartoon.................................... 79
Celestial Map..................................... 82
Summary.............................................. 86
5. DISCUSSION........................................... 87
The Rhetorical Context............................... 88
ParticleWhat It Is............................... 88
WaveWhat It Does................................. 88
FieldHow It Works................................ 89
The Graphics of Spinning Comets.................... 91
Step 1. Comets in General......................... 91
n


Step 2. Focus on the Nucleus...................... 95
Comparing Two Usages.............................. 97
Step 3. Explanation of Observations............... 103
Redesigning a Graph............................... 106
Step 4. Physical Forces........................... 109
Directing Attention within a Graphic.............. 109
Realigning a Wobbly Metaphor...................... 113
Step 5. Results and Predictions................... 114
Summary.............................................. 118
Conclusions and Questions............................ 120
APPENDIX A: "SPINNING COMETS"........................... 122
APPENDIX B: EXCERPT FROM "THE ROTATION OF COMET
NUCLEI"............................................. 134
REFERENCES.............................................. 136
iii


FIGURES
Figure 1. Rodmans continuum of abstraction............... 14
Figure 2. Grid of perspectives............................ 18
Figure 3. Grid for tagmemic analyses...................... 19
Figure 4a. Brownlee and Kirtzs bar graph analysis......... 21
Figure 4b. Brownlee and Kirtzs line graph analysis........ 22
Figure 5. Funding table.................................... 24
Figure 6. Funding bar graph................................ 25
Figure 7. Funding table and bar graph questions........... 26
Figure 8. Coggias comet of 1874 on July 13 (drawn by
Brodie)................................................. 50
Figure 9. Analysis grid of Brodie's drawing............... 51
Figure 10. Abstraction continua with Brodie's drawing..... 52
Figure 11. Tebbutt's great comet of 1861 (drawing by J. F.
Julius Schmidt, Vienna)................................. 53
Figure 12. Analysis grid of Tebbutt's comet................ 54
Figure 13. Abstraction continua with Tebbutt's comet...... 55
Figure 14. Donatis comet of 1858 (drawn by Otto Struve at
Pulkovo on October 5)................................... 57
Figure 15. Analysis grid of Struves drawing.............. 58
Figure 16. Abstraction continua with Struve's drawing..... 59
Figure 17. Donatis comet of 1858......................... 60
IV


Figure 18. Analysis grid of Donati's comet................. 61
Figure 19. Abstraction continua with Donatis comet........ 62
Figure 20. P/Schwassmann-Wachmann 1........................ 64
Figure 21. Analysis grid of P/SW1.......................... 65
Figure 22. Abstraction continua with P/SW1................. 66
Figure 23. Zdenek Sekanina................................. 68
Figure 24. Analysis grid of Sekaninas photo............... 69
Figure 25. Abstraction continua with Sekaninas photo.... 70
Figure 26. Graph of Enckes comet.......................... 72
Figure 27. Analysis grid of Enckes graph.................. 73
Figure 28. Abstraction continua with Enckes graph......... 74
Figure 29. Jet force diagram............................... 76
Figure 30. Analysis grid of the jet force diagram.......... 77
Figure 31. Abstraction continua with the jet force diagram... 78
Figure 32. The wobble of precession........................ 79
Figure 33. Analysis grid of the wobble cartoon............. 80
Figure 34. Abstraction continua with the wobble cartoon.. 81
Figure 35. Celestial map................................... 83
Figure 36. Analysis grid of the celestial map.............. 84
Figure 37. Abstraction continua with the celestial map... 85
Figure 38. Comets in general............................... 92
Figure 39. Focus on the nucleus............................ 95
Figure 40. Graphics from the "Rotation" chapter............ 96
v


Figure 41. Explanation of observations.................... 104
Figure 42. Sketches of Encke's graph...................... 107
Figure 43. Physical forces................................ 110
Figure 44. Sketch of a redesigned jet force diagram....... 112
Figure 45. Sketch of a redesigned wobble cartoon.......... 114
Figure 46. Results and predictions........................ 116
vi


TABLES
Table 1. Dondis Continua of Visual Elements.......... 11
Table 2. McKims Dichotomy of Abstraction vs Reality.. 13
Table 3. Chapter titles from The Mystery of Comets.... 42
Table 4. Computer-generated text analysis............. 43
Table 5. Captions and graphics from Spinning Comets. 46
Table 6. Comparison of the two texts.................. 100
Vll


ACKNOWLEDGEMENTS
This thesis culminates a graduate program that has enriched
my professional life and introduced me to uncountable new ideas.
Many people have contributed to this experience. My fellow students
and my instructors have been interesting and stimulating company.
My advisor, Charles Beck, has encouraged and supported my
efforts through several classes and throughout the preparation of this
thesis.
Scott Grabinger and Daniel Montano graciously spared time to
serve on my committee. Their comments gave me much-needed
insight.
Professor Fred L. Whipples book, The Mystery of Comets, gave
me an example of well-written communication between an expert and
a lay audience. I thank Professor Whipple for allowing me to dissect
his work and the Smithsonian Institute and the University of Arizona
Press for permitting me to reproduce that work.
Charles Barth and the Laboratory for Atmospheric and Space
Physics of the University of Colorado at Boulder have accommodated
yet another graduate student.
Throughout it all, Alain Jouchoux met my bus late at night,
fixed dinner, and took me and the dog for walks when we needed to get
away. He always reads the pictures.
Vlll


1. INTRODUCTION
The invention of the camera has brought about a
dramatic new view of communication and, collaterally,
of education . visual media not yet in currency will
modify our definition not only of education but also of
intelligence itself.. An area that was once the
exclusive province of the artist and designer must now
be considered the concern of both those who work in any
of the the visual media and their audience.
Donis A. Dondis
This thesis broaches the topic of designing illustrative material,
or graphics, to accompany technical material. Technical
communicators are faced with the problem of designing and using
graphics effectively every day, whether they work in the traditional
print media, with on-line documentation, with still photography, or
with videography.
Most writers and editors value illustrative material highly. We
instinctively agree that "a picture is worth a thousand words." We are
inundated with advice on fabricating and reproducing that picture;
however, most of us have very few resources to inform us about
designing illustrations and even fewer to indicate how to use them to
the best advantage. This thesis opens an exploration of how technical
communicators can use graphics to convey our rhetorical message.
Among the aspects of visuals open for discussion, the level of
abstraction of illustrative material can serve as a point of departure for
discussing how text and graphics work. We can analyze the level of
abstraction of both the form and the content of a graphic and use the
insight gained to integrate the graphics and the text. As skilled


communicators, we should be able to address these aspects of graphic
design and to employ them to best advantage.
We face choices everyday in using a photograph of an object, a
sketch of the object, or a schematic diagram of the system in which the
object fits. This thesis is an attempt to find a way to discuss these
choices and how they affect our message.
The Problem of Designing Effective Graphics
Graphics (reproductions of drawings, photographs, maps,
diagrams, or even tables) serve as valuable tools to technical
communicators, although few studies guide us in designing effective
graphics. Efforts to impose rigid rule systems to define, describe, and
use graphics can lead us to neglect the goal of graphics in technical
publications, effective communication. If we are to use these powerful
tools effectively, technical communicators will have to consider the
communication situation and the intentions of the designer, among
many other elements. We must, then, look for ways to analyze the
effectiveness of graphics in the context of the rhetorical situation.
Producing effective graphics is a complex enterprise whose
success depends on several independent factors. One critical factor is
execution, since the most exact drawing is unintelligible if badly
reproduced (Felker 1980). In a study of visual literacy, Donis A.
Dondis (1973) stressed the impossibility of producing art (including
technical artwork) without craft. Another factor, poor presentation,
also attenuates the impact of a graphic display we have all puzzled
over text that refers to non-existent figures, over illustrations that do
not match their descriptions in the captions, or over numbers and
letters rendered illegible from excess photocopying and build-up of
2


correction fluid. We can correct these sorts of problems, usually at the
cost of a little more time for editing and reproduction.
More subtle problems arise in the design of graphics for a
particular piece of communication. Designers of graphics, like
designers of anything else, must make two basic decisions: what
information to include and how to present that information. When we
make these decisions in light of audience analysis and rhetorical
strategies adopted for the specific situation, we approach the goal of
effective communication.
We make these two basic decisions, what to say and how to say
it, along with a series of collateral decisions:
when are graphics needed?
what part of the message do the graphics convey?
how do we relate the graphics message to the content of the
text?
what form should the graphics take?
how do we integrate the graphics into the body of the document?
how will the document be reproduced?
how well does the graphic display fit the medium?
All of these decisions and choices form the context in which we use
graphics.
Our texts and handbooks often deal with production techniques
and aesthetic considerations. For example, D. H. Cunningham and
G. Cohen (1984, 80-81) advise us to use pictures to emphasize only the
subject and to eliminate extraneous details. K. W. Houp and T. E.
Pearsall (1980, 232-253) suggest that visual aids should be used if our
ideas are primarily quantitative, structural, or pictorial. They further
advise us to consider where our skills are in words or visuals
and to concentrate our efforts accordingly.
3


While all of this advice is sound, not much of it rests on
theoretical ground. Alan Gross (1983) and Alan Manning (1989)
sought to study graphics in terms of parallels with language.
Recently, however, Gross (1990) speculated that graphic displays are
composed of such disparate elements that they not likely to be
described according to any theory.
Theoretical analyses and research reported in the literature
concern, mainly, peripheral issues of perception and memory.
Empirical research that deals with graphics has often concerned the
issues of readability and legibility (for example, Macdonald-Ross 1978,
cited by Felker 1980). Much research has focussed on questions of how
humans see, perceive, and remember pictorial information (for
example, Baggett 1984; Canelos, Taylor, and Dwyer 1985; Carpenter
and Just 1976; Eijkman 1984; Fisher 1976; Hochberg 1962; Loftus, 1976;
Loftus and Bell 1975). Specific disciplines have developed guidelines
for the preparation of graphics for specific purposes (for example,
Dickinson 1973; Funkhauser 1938; Tufte 1983,1990; Zeisel, 1985),
especially for the design of computer screens (Kerr 1986; Nickerson,
1985; Norrish 1984; Pullinger 1984; Reed 1985; Rubens, 1988; White and
others 1984). However, narrowly focussed guidelines resulting from
such studies do not generalize easily to other situations.
Some studies have compared how well various subjects recall
information presented in different formats (for example, Powers and
others 1984; Rubens and Rubens 1985,1988). W. Howard Levie and
Richard Lentz (1982) have reviewed a large body of literature that
compares learning from illustrated text and learning from text alone.
They concluded that illustrations, used in certain circumstances, can
facilitate learning. However, conflicting results of various types of
experiments gave no reliable information on how graphics function in
an instructional context. A later study by Jeffrey Hurt (1987) identified
functional roles for illustrations in instructional material that depend
4


in part on their relationship to the text. However, he has yet to explore
this topic in depth.
Seeking Solutions in Rhetorical Studies
One possible way to unify this body of guidelines, advice, and
disparate research results is to examine graphics as rhetorical
elements of communication situations (Gross 1990; Trammel 1988).
Rhetorical analyses of texts that have focussed on audience have
provided powerful insights into the rhetorical situation and the
strategies that effectively communicate information to particular
audiences (Fahnestock 1986; Fahnestock and Secor 1988; Journet 1986;
Rowan 1989; Walzer 1985). Such analyses have demonstrated that text
and strategy change as the audience changes. Consequently, as a
piece of written discourse evolves, we should expect the accompanying
graphics to play different roles, if they are to be effective conveyors of
information.
In this thesis, I will explore the use of graphics as rhetorical
elements. To do so, we must be able not only to describe the
components of a graphic but to discern how those components interact
with each other and with the text. When we address the questions of
graphics, another complication arises: both the form and the content
should be considered as separate, though interacting, tools of
communication. I propose to discuss these two factors of graphic
design in terms of abstraction, since both form and content of graphics
can vary from concrete to abstract (Burton 1985; Gibson 1979; Hale
1972).
At the same time, we must be able to describe the text in a way
that elucidates the role of the graphic. To do so, we will discuss the
communicative intentions of the text in terms of classical rhetoric, in
5


this case, stasis theory. Stasis theory, which was used to develop legal
arguments, points to the underlying issues of a discussion. While
particular techniques may be used to advance or counter an
argument, emphasizing fundamental issues allows us to discuss the
dynamics of the text and its interaction with the graphics.
Combining a discussion of abstraction in graphics and the
rhetorical context requires an orderly process. For this thesis, I have
elected to use a heuristic analysis procedure that specifies viewpoints
to be investigated and orders the questions to be asked. Tagmemic
analysis, developed from linguistic studies, allows us to map the
analysis on a grid or matrix. Through the grids, we can retrace the
analysis of graphics and text. Heuristics, by definition, are not rigid,
but the clarity offered by the tagmemic grids will encourage us to look
at graphics in a consistent, systematic way.
Organization of this Thesis
In these pages we will explore how we can use graphics to serve
our rhetorical goals effectively. This chapter has introduced the
problems that designers have in making choices about the content of
graphic material and the presentation of that content. By considering
the abstraction of graphics and the dynamics of the rhetorical
situation, we can discuss how the graphics work and how to make
them more effective.
In Chapter 2 we will examine a working definition of
abstraction and describe how the concept of abstraction can be depicted
on continua.
Chapter 3 describes how tagmemic analysis techniques apply to
this problem and how they provide an organization for this study. A
6


sample analysis of a table and a bar graph shows how the method
works.
Chapter 4 presents an example of a text taken from popular
scientific literature. I have chosen a chapter from The Mystery of
Comets by Fred L. Whipple. This example is a well-written, concise
treatment of a specialized subject, the rotation of comets, geared
toward a lay audience. We can describe the text in terms of a
computer-generated readability analysis and in terms of classical
stasis theory. Having described the rhetorical context, we can then
describe the graphics in terms of their levels of abstraction.
Chapter 5 discusses the analyses in detail, especially how the
graphics interact with the text. Based on the discussion, I have
proposed some revisions to some of the graphics. The chapter also
poses some questions for further exploration.
7


2. THE QUESTION OF ABSTRACTION
A picture is a surface that always specifies something
other than what it is it preserves what its creator has
noticed and considers worth noticing
J. J. Gibson
We can consider abstraction as a process of expression that
seeks to reveal underlying structures and basic relationships. In the
visual arts, the movement toward abstraction entails manipulating
visual elements to expose essential forms or, as Henri Matisse (1972)
said, to strip away the anecdotal. Another aspect of abstraction that
concerns us here touches on the intellectual processes necessary for
humans to understand and elaborate concepts and ideas in language.
Let us frame this discussion in two parts: abstraction of visual
elements, and abstraction of language elements. This division
represents, respectively, concerns of form and of content of visual
communication.
We will discuss form and content of graphics in terms of syntax
and semantics. This analogy comes from discussions of visual
literacy (Ausburn and Ausbum 1978; Dondis 1973). While the visual
literacy movement in educational philosophy may not have universal
approval (see, for example, Cassidy and Knowlton (1983) and Sless
(1984)), it does provide us with a useful framework to address problems
in the design of graphics.
Addressing the question of syntax of graphics, Dondis (1973)
made the following distinction between linguistic and visual literacy:
In language syntax means the orderly arrangement of
words in their appropriate form and order. ... But
syntax in the context of visual literacy can only mean the


orderly arrangements of parts, leaving us with the
problem of how we can approach the process of
composition with intelligence and knowledge of how
compositional decisions will affect the final result. (19)
In this thesis, I will use the term visual literacy to describe the
abilities of human beings to understand pictorial messages. The term
syntax will designate the patterns of pictorial elements that form a
graphic. The term semantics will designate the interactions of the
graphic and the text that convey the message.
Visual Elements
Dondis (1973) enumerates the parts of a graphic, or the visual
elements, from an artistic viewpoint:
dot the minimal visual unit pointer,
texture
tone
color
direction
line
shape
marker of space
the fluid articulator of form
the basic forms circle, square,
triangle, and their planal and
dimensional variations
the thrust of movement circular,
diagonal, and perpendicular
the presence or absence of light
the coordinate of tone with the added
component of chroma
the surface character of visual
materials, optical or tactile
scale or
proportion relative size and measurement
9


dimension
and motion magnitude and change, both as
frequently implied as expressed
The arrangements of these visual elements make up the syntax of
graphics. We may manipulate the visual elements in various ways to
express the desired message. Gestalt theories of perception describe
how we see and interpret such information (Bernhardt 1986; Ellis
1938; Wertheimer 1938). For example, to make something look larger,
we can place it next to something small; or, to draw attention to
something, we can place it out of alignment or isolate it. According to
Dondis, some techniques to manipulate the visual elements extend
along various continua, as shown in Table 1.
If we consider the arrangement of the visual elements as
variations along these continua, we can modulate our analyses of
graphics very precisely. Dondis formulated the continua in Table 1 to
describe various artistic techniques used to achieve the syntactical
arrangement of a graphic. We must now extend the discussion to
include the semantic aspects of graphics.
The Concrete-Abstract Continuum
Robert McKim (1972) asserted that the dimension of abstract-to-
concrete pervades all cognitive activity, a view that also echoes in the
studies of visual literacy and perception of Dondis (1973), Gibson (1979),
and N. C. Hale (1972). Language, one of the primary cognitive
activities of our species, is commonly analyzed in terms of abstraction
and concretization, notably by S. I. Hayakawa (1964; who bases much
of his work on earlier studies by Alfred Korzybski, especially referring
10


Table 1. Dondis Continua of Visual Elements
Some common techniques to vary visual elements
Contrast Harmony
Exaggeration Understatement
Spontaneity Predictability
Accent Neutrality
Instability Balance
Fragmentation Unity
Economy Intricacy
Boldness Subtlety
Transparency Opacity
Variation Consistency
Complexity Simplicity
Distortion Realism
Depth Flatness
Sharpness Diffusion
Activeness Passiveness
Randomness Sequentiality
Irregularity Regularity
Juxtaposition Singularity
Angularity Roundness
Representation Abstraction
Verticality Horizontality
to Science and Sanity: An Introduction to Non-Aristotelian Systems
and General Semantics).
Hayakawa (1964,179) constructed a hierarchy, or ladder, of
levels of abstraction in the comprehension and use of language. Using
the example of Bessie the cow, he made the point that reality cannot be
11


completely described by words. Bessie, he says, is a living organism,
a whirl of electro-chemico-neural eventfulness, which, like the
universe around her, is in a perpetual state of flux. Presented with
Bessie, we begin by abstracting:
disregarding the dynamic processes, we perceive an object that
we classify as cow;
we name this particular member of the cow class Bessie;
we can refer to Bessie as a cow, as livestock, as a farm
asset, each time abstracting different sets of information;
and, finally,
if we include Bessie in the term wealth, we have eliminated
almost all reference to Bessie as an animal, or a cow, or an
individual.
Despite certain correspondences between linguistic and visual
literacy, Dondis cautioned that we may not consider visual literacy as
a clear-but-logical system analogous to language. Languages are
artificial systems constructed to encode, store, and decode
information. The structure of language systems, therefore, has a
logic that visual literacy cannot parallel, since visual literacy depends
heavily on the physiological processes of perception.
After qualifying the analogies between language systems and
visual systems, different authors approach the definition of
abstraction differently. Dondis (1973) described visual data as having
three distinctive and individual levels: the visual input, which
consists of symbol systems; the representational visual material we
recognize in the environment and can replicate; and the abstract
understructure.
In devising his definition, McKim (1972) pointed out that we
have a tendency to seek overall visual patterns first and then to
analyze details. He categorized various types of graphics
hierarchically as sets of rules that express more or less abstract ideas.
Table 2 presents McKims system. His abstract graphic languages
12


include graphics that present complex, symbolic relationships. The
concrete graphic languages in Table 2 are mimetic depictions of
individual objects.
Table 2. McKims Dichotomy of Abstraction vs Reality
Abstraction
abstract graphic languages
charts
graphs
diagrams
schematics
Reality
concrete graphic languages
orthographic projections
isometric and oblique projections
perspective projections
three-dimensional mockups
appearance or working models
Lilita Rodman (1985) extended McKims analysis to enable us to
categorize graphics, not only by form but also by content. She advised
teachers of technical communication to discuss graphics in terms of
abstraction, which she describes as continua of form and content.
While a formal definition of abstraction has not yet received general
acceptance, Rodman observed three general tendencies as graphics
become more abstract:
A movement from the depiction of the particular to the depiction of
the generic;
A movement from a focus on surface characteristics to a focus on
structure or organization; and
A movement from mimetic to symbolic means of representation.
These tendencies, Rodman said, constitute a continuum from
concrete to abstract. She conflated a continuum of form with another
of subject: the photograph of a specific object is more concrete than an
13


aerial photograph of a region. Figure 1 presents an adaptation of
Rodmans continuum.
Levels of Abstraction
Generic,
Structural,
Symbolic,
i
diagram
[syntax]
drawing
map
line
graph
bar graph
table
flow
chart
photo
aerial
photo
object
rea quantity process
Individual,
Surface, [semantics]
Mimetic
Figure 1. Rodmans continuum of abstraction
14


While Rodman described a useful means of discussing graphics
with her students, she does not relate abstraction of graphics to the
rhetorical situation. Other authors have examined how varying levels
of detail in graphics affect how well readers retain certain types
information (for example, Bahrick and Boucher 1968; Loftus and Bell
1975; Nelson, Metzler, and Reed 1974; Nelson, Magliaro, and Sherman
1988). However, these studies do not link rhetorical concerns with
graphics, with the level of detail presented, or with other aspects of
abstraction. We can link graphics to text if we examine them as
rhetorical elements.
This chapter has explored the topic of abstraction as a means to
approach the problem of designing effective graphics. The Rodman
grid gives us a continuum along which we may situate various
graphics. Let us now see how we may assign specific graphics to a
position on that grid as a function of their rhetorical goal. To do so, I
will borrow a powerful analytical tool from linguistic theory,
tagmemic analysis.
15


3. TAGMEMICS: A METHODICAL ANALYSIS
... a theory and a technique which would pass without a
jar from the study of the structure of one kind of activity
of man to that of any other kind. Ideally, this would
result in one basic theory of structure, one basic set of
terms, and one basic methodology which could be
applied to the analysis of language, the analysis of ritual
behavior, the analysis of sports, the analysis of
occupational activities, or even to the processes of
thought itself.
Kenneth L. Pike
At a NASA-sponsored symposium on technical communication
(Mathes and Pinelli 1981), Pamela Brownlee and Mary Kirtz
addressed the problem of teaching graphics in technical writing
courses; they suggested using a heuristic approach, specifically,
tagmemic analysis. They described their approach to teaching
technical writing as addressing two fundamental problems: writers
had trouble with invention (what material to include) and with
arrangement (how to present the material). They included graphics
in this discussion because they considered the graphic mode as an
ancillary method of reporting information. They selected an approach
adapted from linguistic theory, tagmemics, a problem-oriented
heuristic procedure that is readily transferable to technical writing
and to the design of graphics for technical publication (Pike 1972,1981,
1982; Pike and Pike 1983; Waterhouse 1974; Young, Becker, and Pike
1970).


Tagmemics: A Translators Tool
Tagmemic theory grew from efforts to analyze unwritten
languages and to translate material into those languages (Brend,
1972; Pike 1981,1982; Pike and Pike 1983; Waterhouse 1974). The
linguist Kenneth L. Pike developed tagmemics through the Summer
Institute of Linguistics and the Wycliffe Bible Translators. Pikes field
work began in 1935 in Mexico where he worked with a group of Mixtec
speakers. Their goal was to enable missionaries to translate the Bible
into the languages of their congregations.
Pike regarded language as human behavior inseparable from
its cultural context. His efforts to analyze language incorporated
context and hierarchies at all levels. Pike redefined a concept
developed earlier in linguistic research, the tagmeme, as a unit-in-
context. He then proceeded to examine this unit-in-context from a
variety of viewpoints in an effort to define not only the function and
position of the unit itself, but its place in a context and its interaction
with its environment.
From various perspectives Pike examined units of grammar in
order to formulate a context for parts of language. From the context,
Pike could then extrapolate grammatical rules which he could test
against experience. These same techniques can be used for
anthropological studies of the customs of different societies (Young,
Becker, and Pike 1970).
Borrowing from the vocabulary of optical physics, Pikes system
encourages us to examine the functions of any entity:
as a particle (an individual unit),
as a wave (a dynamic factor), and
as a field (a system).
17


These functional perspectives (shown in Figure 2) can be further
elaborated from differing points of view: their contrast (what makes
them unique), their variation (how they differ from other things of the
same kind), and their distribution (how they fit into a larger system,
network, or environment).
Contrast Variation Distribution
Particle What it is What it is What it is
What is unique about it as an individual, isolated object? How is it, as an individual, isolated object, different from others of the same kind? Where is it, as an individual, isolated object, found?
Wave What it does What it does What it does
What is unique about it as a dynamic event or object? How is it, as a dynamic event or object, different from others of the same kind? Where is it, as a dynamic event or object, found?
Field How it works How it works How it works
What is unique about it as a system of components with goals? How is it, as a system of components with goals, different from others of the same kind Where is it, as a system of components with goals, found?
Figure 2. Grid of perspectives
18


The process of constructing a tagmemic analysis produces a
somewhat cumbersome, intersecting grid of nine aspects for exploring
the definition, form, function, and context of an entity. The grid in
Figure 3 is adapted from that produced by Young, Becker, and Pike
(1970).
Contrast Variation Distribution
Particle What it is What it is What it is
isolated entity one of a group classified in a larger context
Wave What it does What it does What it does
dynamic event or object dynamic process, change part of a dynamic environment
Field How it works How it works How it works
system of parts or elements system compared to other systems system interacting in an environment
Figure 3. Grid for tagmemic analyses
As with other heuristic procedures, the questions generated by
the tagmemic grid are open-ended and encourage exploration. We
can view the open-endedness of this heuristic process as both a
strength and a weakness. Richard W. Bailey (in Pike 1981, xv)
explained that while tagmemic descriptions have no necessary
starting place, they have a common goal: the description of behavior
within the context of a universe of discourse. In a discussion of
classical invention techniques, E. P. J. Corbett (1986) speculated that
since heuristic devices are thought to mirror the way in which the
mind works, we unconsciously ask the sorts of questions proposed by
the ancient topoi and by modem methods such as neo-classical
19


invention, Kenneth Burke's dramatistic method, D. Gordon Rohman's
pre-writing, and Kenneth Pike's tagmemic analysis. Corbett added
that systematically formalizing a set of questions provides us with a
simple means of monitoring our coverage of the possibilities. The
tagmemic analysis technique that I employ in this thesis allows me to
formalize my questions and to document, in a systematic manner, the
analyses of a set of graphics.
Next, I will recapitulate the reasoning used by Brownlee and
Kirtz in their example of how tagmemic analysis can profitably be
used to study graphics. Then, we can follow a sample analysis to
demonstrate how tagmemic procedures work in individual cases,
what kinds of information they elicit, and how they adapt to graphics.
Examples of Tagmemic Analyses
Bar Graphs vs. Line Graphs
The nine aspects, or viewpoints, displayed in Figure 3 allow us
to consider graphics as
individual pictorial descriptions of information,
dynamic conveyors of information, and
elements in the overall structure.
Brownlee and Kirtz (1981) compared bar graphs to line graphs from
these points of view. Figure 4a and Figure 4b display a recapitulation
of their analysis. They found that the tagmemic heuristic focussed
attention on the relations of parts to the whole. As a result, they could
consider the graphics as working entities instead of simple pictorial
descriptions of data.
20


Contrast Variation Distribution
Particle What it is as an isolated entity Particle What makes it different from other group members Particle What it is in context
on a matrix has X-Y axis individual bars can have any number of bars X-Y axis useful for precision many possible distributions (negative, etc.) compares parts and different wholes emphasizes static presentation
Wave What it does as a dynamic unit Wave What it does differently from other dynamic processes Wave What it does part of dynamic environment
bars indicate dimensions significance of information changes according to type static comparison divided into proportions
Field How it works as a unique system Field How it works differently from other systems Field How it works system within an environment
all parts work to show precise proportions could be temporal, but represents spatial better multiple bars: accentuate parts 100% bar: stresses parts-to-whole sliding bars: stresses negative as well as positive shows dimensions, proportions emphasizes stasis focuses on data, on the parts themselves
Figure 4a. Brownlee and Kirtzs bar graph analysis
21


Contrast Variation Distribution
Particle What it is as an isolated entity Particle What makes it different from other group members Particle What it is in context
on a matrix X-Y axis dots connected linearly can have unlimited number of dots can be differentiated by color or line variations compares sequences emphasizes dynamic presentation
Wave What it does as a dynamic unit Wave What it does differently from other dynamic processes Wave What it does part of dynamic environment
connected dots show variations in dimension, amount, proportion axis can stand for any value, time, space or sequence focus on comparison emphasizes movement of data
Field How it works as a unique system Field How it works differently from other systems Field How it works system within an environment
stresses variation between at least 2 points emphasizes temporal and sequential stresses trends over time stresses cause/effect in sequences shows movement, trends emphasizes dynamics, change, flow rather than each part individually
Figure 4b. Brownlee and Kirtzs line graph analysis
22


The individual bars of bar graphs indicate relationships among
the data elements, dimensions, and proportions. The disposition of
these bars can vary, and each type of bar graph emphasizes some
aspect of the data; for example, sliding bar graphs stress negative
relationships as well as positive ones. Bar graphs compare individual
elements in a static presentation; they focus on individual data, or
individual parts.
The connected dots of line graphs, according to Brownlee and
Kirtzs analysis, show variations in dimension, amount and
proportion, and they emphasize sequential and temporal
relationships. Line graphs may show vast numbers of data points,
may be differentiated by color or line variations, and the axes may
stand for any value, time, space, or sequence. Line graphs stress
trends over time and cause-and-effect relationships in sequences.
They focus on sequences and changes of data and emphasize
dynamics and flow of the whole instead of individual data points.
If we assume, as did Brownlee and Kirtz, that the primary goals
of graphics are
to attract attention, and
to increase understanding through an additional perspective on
material already communicated verbally,
we can begin to address whether these types of graphics meet their
goals. If the intent of the designer/writer is to emphasize changes in
data from one reading to the next, then a bar graph is an appropriate
choice. If the intent, however, is to give a broad overview of the entire
data set and how data change over time, then a line graph is more
appropriate.
23


Analysis of Table vs. Bar Graph
Let us now consider two graphics: a table of funding amounts
and a bar graph of funding data. Figure 5 presents the table of
funding information, and Figure 6 presents a bar-graph version of
Appropriated Funding
(in millions of dollars)
July 1966 July 1988
1967 Dollars Current Dollars 1987 Dollars
1966 0.760 0.739 1.265
1967 1.992 1.992 3.394
1968 2.285 2.388 4.039
1969 1.812 1.997 3.345
1970 1.854 2.166 3.586
1971 1.866 2.273 3.730
1972 2.343 2.941 4.791
1973 2.194 2.911 4.675
1974 2.026 2.998 4.685
1975 1.608 2.609 3.973
1976 1.504 2.573 3.847
1977 0.990 1.808 2.644
1978 1.174 2.320 3.298
1979 1.216 2.662 3.610
1980 1.845 4.572 5.803
1981 1.692 4.644 5.542
1982 1.318 3.857 4.412
1983 1.395 4.176 4.662
1984 2.243 6.990 7.541
1985 1.913 6.145 6.427
1986 2.058 7.617 7.827
1987 1.924 6.498 6.498
1988 7.276
Figure 5. Funding table
24


Appropriated Funding
(in millions of dollars)
July 1980 July 1988
10
9
8
7
6
5
4
3
2
1
0
Figure 6. Funding bar graph
similar information. We will begin with the information presented in
the two graphics and the portrayal of that information in the report.
The information contained in the two graphics was included in
reports of the activities of a research and development organization.
This report fills a number of needs:
a reporting tool to superior administrators;
a recruiting tool sent to prospective employees;
a public relations tool to describe accomplishments;
25
(millions of dollars)


an in-house communication tool for employees;
an internal history of the organization, recording publications,
achievements, and so on.
Figure 7 presents two grids of questions to be answered for both
the table and the bar graph.
Funding table:
Contrast Variation Distribution
Particle What it is as an isolated entity Particle What makes it different from others Particle What it is in context
What are the identifying features? How can the elements vary? How are tables typically used?
What are the composing elements ? Should we use different values? Is this table appropriate?
Are titles and headings proper? Should there be other kinds of information? What kind of data are presented?
Wave What it does as a dynamic unit Wave What it does differently Wave What it does as part of dynamic environment
What are the dynamic aspects? How do the dynamics work? How does the table interact with ?
What kinds of sequences are shown? How do the data relate? What does it tell about the data?
How does the data relate to the labels? How does it carry the message?
Field How it works as a unique system Field How it works differently Field How it works as a system within an environment
How are the components organized? What other forms can the components take? How does it contribute to the publication?
How do components interrelate? How does it integrate into the report?
Figure 7. Funding table and bar graph questions
26


bar graph:
Contrast Variation Distribution
Particle Particle Particle
What it is as an isolated What makes it different What it is
entity from others in context
What are the How can the elements How are bar graphs
identifying features? vary? typically used?
What elements make Would different Is this one appropriate?
up the bar graph? inflation values be of
use? What kind of data are
Are the titles and presented?
headings appropriate? Should there be other kinds of information?
Wave Wave Wave
What it does as a What it does differently What it does as part of
dynamic unit dynamic environment
What are the dynamic How do the dynamics How does it interact?
aspects? work? What does it tell about
What kinds of How do the data the data?
sequences are shown? relate? How does it carry the
How does the data relate to the labels? message?
Field Field Field
How it works as a How it works How it works as a
unique system differently system within an environment
How are the What other forms can How does it contribute
components organized? the components take? to the report?
How do they How does it integrate
interrelate? into the report?
Figure 7. Funding table and bar graph questions (continued)
Because the report is considered a public document, the
particular organizational structure is such that detailed financial
information can be transmitted to administrative officers in other
27


ways. In fact, this report is not a channel for transmitting detailed
information. The financial information included in the report is, by
design, of a broad, general nature. Given this context of publication,
let us now proceed through a tagmemic analysis of the two graphics,
beginning with the funding table (Figure 5) and then the bar graph
(Figure 6).
Funding Table Particle/Contrast
What it is (as an isolated entity)
What are the identifying features?
What elements make up the table?
Are the titles and headings appropriate?
The table in Figure 5 has a heading and four columns of
numbers enclosed in a box.
The first column lists the year for the funding numbers in the
other three columns. The second column lists the inflation-
corrected figures in 1967 dollars. The third column lists
Current Dollars, that is, the actual amounts awarded each
year with no correction for inflation. The third column lists
inflation-corrected figures in 1987 Dollars.
The table reports funding in millions of dollars, and the figures
are carried to the third decimal place.
28


Funding Table Parti cl e/Variation
What it is (that makes it different)
How can the elements vary?
Should there be other kinds of information?
Tables can carry unlimited numbers of columns and rows.
Usually, typographic devices, such as rules between headings,
underlining headings, or highlighting every fifth row, will
increase legibility. Whether the information is displayed as a
column or a row is indifferent that is, there could be 22
columns of four rows each instead of four columns of 22 rows.
Other sorts of inflation values, based on different years, could
provide a context for comparing dollar values.
Funding Table Particle/Distribution
What it is (within its context)
How are tables typically used?
Is this table appropriate to this kind of publication?
What kind of data are presented?
Tables provide precise data values. They should be used to
present raw data since they emphasize no relationships
among the data.
Since degrees of precision can vary, we should determine
whether they are appropriate to the publication. In this case,
since the publication purposely does not include detailed
financial information, when speaking of millions of dollars,
we should consider whether precision to the nearest thousand
dollars is too detailed. Perhaps precision to the nearest ten
29


thousand dollars would retain the same impact without
bogging down into strings of numbers.
These are raw numbers. The text provides no interpretation of
their significance.
Funding Table Wave/Contrast
What it does (as a dynamic unit)
What are the dynamic aspects?
What kinds of sequences are shown?
The tables main purpose is to present data in plain columns and
rows of numbers
The predominant relationship here is temporal, and the visual
aspect remains entirely static.
Funding Table Wave/Variation
What it does (that is different from others)
How do the dynamics work?
How do the data relate?
How do the data relate to the labels and other parts?
The parade of years from 1966 to 1988 gives a sense of the passage
of time.
Some tables give a visual sense of magnitude by clustering data
or forming patterns.
Individual data points can be found easily, and the labels clearly
indicate their categorization. However, relationships among
theses data are not evident.
30


Funding Table Wave/Distribution
What it does (as part of a dynamic environment)
How does the table interact with the other material?
What does it tell about the data?
How does it carry the message?
The financial information within this report remains very
general. The table appears in a section listing current active
grants and programs. Introduced by a short paragraph
announcing the number of programs and the years total
appropriated funding, the table occupies an entire page. The
material before the table is the general introduction and
overview; the pages immediately following carry the list of
program titles, funding sources, dates, and responsible
personnel.
This relative isolation of the funding table reinforces the general,
non-specific nature of this information; it does not relate
directly to any other information in the report.
No specifics are given, and no specific conclusions can be drawn.
Funding Table Field/Contrast
How it works (as a unique system)
How are the components organized?
How do the components interrelate?
The data entries are arranged chronologically. The date column
on the left displays the year index of the funding amounts.
The column of Current Dollars generates the other two columns.
Current Dollars are multiplied by inflation factors to obtain
31


values in 1967 Dollars or 1987 Dollars. The three columns of
funding numbers reiterate the chronological order instead of
the order of calculation.
Funding Table Field/Variation
How it works (differently from other similar systems)
What other forms can the components take?
A more technical, more specific presentation would include the
inflation factors and make the calculation explicit. Other sorts
of information included in such tables would be an index of
change from year to year. We often see some sort of
highlighting or variable spacing for legibility in such tables.
Unusually high or low numbers can sometimes appear
highlighted. Groupings according to source could also add
information.
Funding Table Field/Distribution
How it works (as a system within an environment)
What identifies this table as part of the publication?
How does it contribute to the report?
How does it integrate into the report?
Within the publication, the table appears on a page with a header
and footer. The identity of the organization is indicated in the
main heading.
An indication of financial standing traditionally forms a part of
such reports. This table provides a general view of the
institutions finances.
32


The table appears in a section that lists currently active grants
and programs.
Funding Table Conclusions
To summarize, throughout this process of analyzing the
funding table, we have seen that the presentation of data without
comment or clarification reinforces the inherently static nature of the
table.
Let us now turn to the bar graph of Figure 6.
Funding Bar Graph Particle/Contrast
What it is (as an isolated entity)
What are the identifying features?
What elements make up the bar graph?
Are the titles and headings appropriate?
The bar graph of funding presents information only from 1980 on.
For each year, two values are reported the amount awarded in
Current Dollars, in white bars, and the value corrected for
inflation in 1988 Dollars, in shaded bars. The bar graph sits
on a square grid. The years form the x axis; the amount in
millions of dollars, the y axis.
The titles, headings and labels are accurate.
33


Funding Bar Graph Particle/Variation
What it is (that makes it different from others)
How can the elements vary?
Should there be other kinds of information?
Instead of a gridded box, other graphs form axes with simple
lines. In fact, Tuft (1983) encourages us to eliminate all excess
ink.
Within the limits of legibility, we can add information by using
different shadings and colors. The table in Figure 5 listed
inflation corrections in 1967 Dollars, and we could add another
bar in each year without making the graph illegible. However,
while we can easily accept that there is a relationship between
Current Dollars and 1988 Dollars as displayed in Figure 6, a
three-way relationship would not be so evident without
explanation.
Funding Bar Graph Particle/Distribution
What it is (in its context)
How are bar graphs typically used?
Is this bar graph appropriate to this publication?
What kinds of data are presented?
Bar graphs typically display whole units of data, instead of
individual data points. The discrete bars encourage us to
compare the two entries for each year or to compare one year
to another instead of comparing funding levels across several
years.
34


The bar graph presents very general information, an inflation
correction and a not-too-precise dollar figure. Because specific
information is not conveyed in this report, the level of detail
here is probably appropriate
We cannot determine precisely the amounts awarded. We can
only estimate within half-million units.
Funding Bar Graph Wave/Contrast
What it does (as a dynamic unit)
What are the dynamic aspects?
What kinds of sequences are shown?
The bars dimensions emphasize the relative importance of the
data they represent. Each year carries two sorts of bars, and
their proximity and the repetition of the two sorts of bars imply
a relationship between the two data sets. If there were only
one bar per year, we could more easily compare magnitude
from year to year.
The temporal sequence is the most apparent. The portrayal of the
inflation correction implies a correlation from one bar to
another which may not exist.
35


Funding Bar Graph Wave/Variation
What it does (differently from other dynamic processes)
How do the dynamics work?
How do the data relate?
How do the data relate to the labels?
Instead of the precision to three decimal points that we found in
the funding table of Figure 5, the y axis marks round numbers
of millions of dollars.
The dimensions of the bars indicate the degree of increase or
decrease of funding. Some bar graphs do specify the exact
values they represent. By diminishing the precision of the
funding amount, we shift the emphasis from the actual
amount and encourage comparisons of magnitude.
The bars representing Current Dollars and 1988 dollars receive
most of our attention, and the comparison between them takes
on greater importance. The correction for inflation serves not
as a simple clarification for the reader, but a major thrust of
the graphic. The unusual position of the y axis further
reinforces this lack of emphasis on the funding amount.
Gestalt theory tells us that our first instinct in looking at a
graphic is to establish the vertical axis as the eye travels
downward, we see the x axis of years with no label. The
second instinct leads us to situate the horizontal axis (Zusne
1970; Dondis 1973; Bernhardt 1986). Reading from left to right,
we first find the key to the shading of the bars, and then we
find the y-axis labels that tell us we are to compare funding
amounts over time.
36


Funding Bar Graph Wave/Distribution
What it does (as part of a dynamic environment)
How does the bar graph interact with other material?
What does it tell about the data?
How does it carry the message?
The bar graph follows the introduction and overview section of
the report. A short paragraph introduces the graphic and the
list of currently active programs.
The data being compared are funding amounts over time, but
these axes are either not labelled or are placed so that they
seem inconsequential.
The discrete bars chunk the data into segments containing two
bars each. The overall pattern encourages us to compare the
white bar and the shaded bar in each segment or to compare
separate segments.
Funding Bar Graph Field/Contrast
How it works (as a unique system)
How are the components organized?
How do the components interrelate?
Each white bar has a companion shaded bar, and our tendency is
to compare the differences between the two.
The gridded box serves to reunite the separate segments and
gives us a measure of the magnitude of the bars.
37


Funding Bar Graph Field/Variation
How it works (differently from other systems)
What other forms can the components take?
The y axis most commonly is labelled on the left side, and the x
axis usually carries a typographically-matching label.
Parallel labelling of the axes would satisfy a Gestalt
requirement for similarity (Zusne 1970). The grid lines are
probably unnecessary (Tufte 1983), although the box does serve
to unify the separate segments.
Funding Bar Graph Field/Distribution
How it works (as a system within an environment)
What identifies this bar graph as part of the report?
How does it contribute to the report?
How does it integrate into the report?
The bar graph of Figure 6 occupies the same place as the funding
table of Figure 5. Its page carries header and footer
information identifying the report.
It adds supplemental information, but this information does not
interweave with other text or graphics.
The isolation of the financial information is intentional and, as
long as no relationships are described by the text or by other
graphics, unavoidable.
38


Funding Bar Graph Conclusions
The bar graph of Figure 6 emphasizes individual segments of
data and encourages us to compare the inflation-corrected values with
the amounts in Current Dollars. The primary relationship was
intended to be that of funding over time; however, by segmenting the
data and omitting the graphic liaisons between the x and y axes, this
graphic blurs the primary relationship and emphasizes the inflation
correction.
Summary
Both types of graphics shown in Figures 5 and 6 have strengths
and weaknesses, and writers/designers may choose which type best
serves their rhetorical goals. The precision of the table allows it to be
an extremely useful tool if the audience needs exact data values.
However, its almost totally static presentation shows no relationships
among the data and a minimum of ranking and ordering. The bar
graph loses precision but does provide a comparison of magnitudes of
change from one segment to another. Its primary focus is on
comparison of units of data, and it may not be suitable for depicting an
overall changing relationship over time.
This chapter has introduced the technique of tagmemic analysis
as a tool for studying graphics as rhetorical elements. We have traced
the background of tagmemics in linguistics and of its application to
the problem of graphic design. A demonstration of the technique has
given us an orderly, informative study of a set of graphics. Let us now
turn to using this technique to assign graphics to the continuum of
abstraction.
39


4. ANALYZING GRAPHICS FROM POPULAR SCIENTIFIC
LITERATURE
... if we try to discover what the poem is doing for the
poet, we may discover a set of generalizations as to what
poems do for everybody.
Kenneth Burke
For our discussion of analyzing graphics for technical
publications, I have chosen one of the chapters from a general-interest
book that Professor Fred L. Whipple published at the time of Comet
Halleys passage: The Mystery of Comets (1985, Washington DC:
Smithsonian Press). Professor Whipple of the Smithsonian
Observatory is one of the worlds authorities on comets. He has
authored papers, articles, and books on the subject for every audience,
from lay people to his peers. I selected this chapter because it presents
a clear discussion aimed toward an informed, but not expert,
audience. Professor Whipple has written on this specific subject, the
rotation of the nuclei of comets, several times for various audiences,
and I will compare the use of similar graphics in different
publications.
To describe the context of a graphic, we must consider the entire
rhetorical situation. The decisions that the author, designers, and
publishers make about the production, size, and cost of a publication
indicate the audience they address. The level of vocabulary, the style of
language, and the organization of the text result from decisions that
constitute the rhetorical situation. I will first describe the text and its
graphics. A computer-generated analysis of the readability of the text
will be used to describe some of the uses of language. After describing


the text on the mechanical level, we will examine the graphics in
terms of their relative abstraction.
Spinning Comets bv Fred L. Whipple
The Mystery of Comets is softbound, has 276 pages including the
index, and measures 6 inches by 9 inches. Its price of under $15.00
makes it economically attractive to the general public. Nearly every
two-page spread carries black-and-white photos and drawings, and an
eight-page color photo section occupies the center of the book.
Table 3 lists the chapter titles. The 24 chapters narrate the
historical background and popular mythology about comets, examine
Halleys comet in particular, recount efforts to discover and study
other comets, describe the current state of humankinds knowledge of
comets, and advocate additional studies and a space mission to a
comet. The first 14 chapters are devoted to pre-20th century history
and theories. The remaining ten chapters concentrate on
humankinds present knowledge and observational strategies.
Chapter 16, Spinning Comets, occupies ten pages and is illustrated
with ten photos, drawings, and graphs. Appendix A reproduces
Spinning Comets.
A Mechanical Analysis
I analyzed the entire text of Spinning Comets with a computer
program, Grammatik (a trademark of Wang Laboratories, Inc.;
copyright 1990 Reference Software International). The program
counts the number of words, sentences, and paragraphs. It
recognizes syllables, prepositions, and passive constructions.
41


Table 3. Chapter titles from The Mystery of Comets
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
Chapter 11
Chapter 12
Chapter 13
Chapter 14
Chapter 15
Chapter 16
Chapter 17
Chapter 18
Chapter 19
Chapter 20
Chapter 21
Chapter 22
Chapter 23
Chapter 24
Forward
Preface
The Fear of Comets
Early Theories of Comets
Breaking the Crystalline Spheres
Halley and His Comet
The Returns of Halleys Comet
The Sport of Comet Hunting
The Consequences of Comet Hunting
Some Comet Frailties and Idiosyncrasies
Small Pieces of Comets
The Magic of the Rainbow
The Heads and Tails of Comets
Sunlight and Comets Tails
The Prodigal Sun
Flying Sandbanks
Dirty Snowballs
Spinning Comets
Comet Landscapes
Comets in the Space Age
The Elemental Composition of Comets
Exotic Ices in Comets
How to Make Comets
Comets and Life on Earth
Comets May Be Dangerous to Your Health
Space Missions to Comets
Epilogue
Additional Readings
Index
42


Using these data, the program also provides three indicators of
readability: the Flesch Reading Ease Score, Gunnings Fog index, and
the Flesch-Kincaid Grade Level. Table 4 presents the summary of the
text analysis.
Table 4. Computer-generated text analysis
Readability Statistics
Flesch Reading Ease: 53
Gunning's Fog Index: 14
Flesch-Kincaid Grade Level: 11
Paragraph Statistics
Number of paragraphs: 18
Average length: 6.7 sentences
Sentence Statistics
Number of sentences: 121
Average length: 20.1 words
End with ?: 6
End with !: 1
Passive voice: 16
Short (< 12 words): 23
Long (> 28 words): 24
Word Statistics
Number of words: 2444
Prepositions: 346
Average length: 4.74 letters
Syllables per word: 1.57
The Flesch Reading Ease Score indicates relative ease of
reading: on a scale from 0 to 100, a low score indicates high difficulty.
The score of 53 rates in the Fairly Difficult category with a
recommendation that some high-school-level education would be
necessary to read the document easily. The formula for calculating
Flesch Reading Ease Score is
43


1.015 x (average sentence length)
+ 0.846 x (number of syllables per 100 words)
= Total
206.835 Total = Flesch Reading Ease Score
Gunnings Fog Index measures approximate grade level
necessary to read easily and understand the document. The score of
14 indicates that a fairly high grade level is necessary for easy
reading. The formula for calculating Gunnings Fog Index is
(average number of words per sentence)
+ (number of words of 3 syllables or more)
= Total
Total x 0.4 = Fog Index
The Flesch-Kincaid Grade Level also indicates grade level. For
general-interest documents, the developers of the program
recommend a grade level between 6th and 10th. A higher score
indicates that the writing might be difficult to understand. Spinning
Comets scored slightly above the recommended difficulty, 11th grade.
The formula used to calculate the Flesch-Kincaid Grade Level is
(0.39) x (average number of words per sentence)
+ (11.8) x (average number of syllables per word)
= Total
Total -15.59 = Flesch-Kincaid Grade Level
These scores (Flesch Reading Ease: 53; Gunning's Fog
Index: 14; Flesch-Kincaid Grade Level: 11) indicate that the difficulty
of reading the text should not be insurmountable to an interested,
adult amateur. The number of short (less than 12 words) and long
(more than 28 words) sentences are nearly equal, and out of a total of
121 sentences, only 16 are in the passive voice.
44


Inventory of the Graphics
Ten figures with captions illustrate the chapter. Table 5 lists
the captions and describes the graphics of Spinning Comets.
Let us now look at the rhetorical roles of the text and the
graphics. We will frame this discussion in the context of classical
stasis theory.
An Expert Speaks to Lav People
Spinning Comets addresses an audience not necessarily
familiar with scientific enquiry or with the study of comets. In the
context of classical stasis theory (Fahnestock 1986; Fahnestock and
Secor 1988; Nadeau 1964), such an audience must be persuaded of the
essential basics of a subject. Stasis theory, a classical rhetorical
invention technique, was developed in the second century B.C. by
Hermagoras of Temnos and was included in Cicero's De Inventione
and De Oratore and in Quintilian's Institutio Oratoria. In the second
century A.D. Hermogenes of Tarsus devoted to stasis theory an entire
volume in his five-volume treatise on rhetoric. Hermogenes,
concentrating on invention in legal argumentation, defined four
issues to be addressed at various points in the formulation of an
argument:
fact what happened and who did it?
Hermogenes used the example of a man burying a newly slain
body in a deserted place. He was discovered and accused of
murder (Nadeau 1964; 390); we can frame a defense in the first
stasis by proving that the dead body was not murdered or that
someone else did it.
45


Table 5. Captions and graphics from Spinning Comets
Caption Description
Coggias comet of 1874 on July 13. (Drawn by Brodie.) drawing
Tebbutts great comet of 1861. (Drawing by J. F. Julius Schmidt, Vienna.) drawing with measurement scale
Donatis comet of 1858. (Drawn by Otto Struve at Pulkovo on October 5.) drawing with directions marked
Donatis comet of 1858. Left, October 6; right, October 7. (Drawn by J. F. Julius Schmidt, Vienna.) 2 small drawings for comparison
Sequence showing an outburst of P/Schwassmann- photographs with dates
Wachmann 1, as photographed on February 1, 10, indicated in the comers
13, and 28, 1981, by R. E. McCrosky and C.-Y.
Shao at Oak Ridge Observatory with the 1.5-meter
reflector, operated by a grant from NASA.
Compare this outburst with that of the same comet
in 1961, shown in Chapter 20.
Zdenek Sekanina portrait photograph
The observed reduction in the change of period of graph plotting year and
Enckes comet since its discovery. The curve gives change in period per orbital
the loss in period for each revolution of 3.3 years revolution
For the oblate comet nucleus in the diagram, the jet force is directed below the center of gravity of the nucleus and so tends to tip the pole counterclockwise, causing precession. cartoon-like drawing with arrows indicating the directions of sunlight and ejected gas and with selected parts labelled.
The wobble of precession. Left, as the top tries to fall over, its pole wobbles in the same sense as its rotation; right, as the Moon and Sun try to tip the pole perpendicular to its orbital plane, the pole left side: cartoon-like drawing of a top with motion indicated by arrows.
precesses in the opposite sense right side: depiction of geometric schema; directions and motions indicated by arrows, angles marked in degrees, labels
Motion of spin of pole of Enckes comet across the gridded celestial map with
sky. Solid curve represents period of observations, latitudes and longitudes
(By Fred L. Whipple and Zdenek Sekanina.) indicated, and the path of the comet through time labelled.
46


definition what was the nature of the act?
Hermogenes used the example of Demosthenes who was a
father, an orator, an ambassador, and a soldier (Nadeau 1964;
389-390); depending on the context of the argument, one must
characterize Demosthenes as the situation dictates,
quality or value what were the circumstances?
Hermogenes used the example of a wartime situation (Nadeau
1964; 392). A tyrant was under siege in his citadel. No one
could find a way into the citadel until a wife disclosed the
secret to her husband. He then entered the citadel, killed the
enemy, and accused his wife of adultery. Hermogenes
commented, "I wonder if anyone will vote against her, even if
she is shown to have committed adultery, for it is through her
that the tyrant has been destroyed."
procedure and policywhat should be done?
Hermogenes used the example of a debate by the Assembly
(Nadeau 1964; 393). The Athenians deliberated on whether
they should honor with burial services those barbarians who
fell at Marathon.
Applying stasis theory to scientific literature, Fahnestock (1986)
and Fahnestock and Secor (1988) added another juncture of
argumentation:
questions of cause what produced the phenomenon?
We will use these categories of Hermogenes and Fahnestock and Secor
to discuss the text:
1st stasis questions of fact
2nd stasis questions of cause
3rd stasis questions of cause
4th stasis questions of value
5th stasis questions of procedure and policy
47


In the first two paragraphs of Chapter 16 "Spinning Comets,"
Whipple explains his thought processes to trace the development of his
theory. He also provides a definite reason for studying comets:
The purpose of studying comets is to help reconstruct these
primitive conditions, which were probably prevalent at the
time when our Sun, Earth, and entire Solar System were
being assembled. Perhaps comets can tell us something
about those ancient times, when the atoms now in our
bodies were floating around in a cosmic cloud. (157)
The first two paragraphs deal with 4th stasis questions of value, an
important topic for a lay audicence. An audience of peers, on the other
hand, would not need to be convinced of the value of pursuing comet
studies in general or trying to determine the cause of certain
unexplained observations.
Overall, paragraphs 3 through 18 address 3rd stasis questions of
cause as they give the historical background of halo studies and trace
Whipples theoretical work. They also treat 2nd stasis questions of
definition as they explain what halos are and how they behave.
The next to last paragraph turns to the 5th stasis question of
procedure and policy: How do we find answers to the scientific
questions? The response: We should launch a space mission to a
comet.
The last paragraph summarizes Whipple and Sekaninas
predictions of how Comet Encke will behave in the future, with some
reservations. Whipple also expresses confidence in the dirty snowball
model.
The ten illustrations are not numbered, nor are they referred to
directly in the text. The first four illustrations are 19th-century
drawings of three comets. Whipple recounts how information has
been collected about comets using full names and giving the location of
the various observers: George P. Bond at Harvard in October 1858, J.F.
Julius Schmidt in Vienna. He explains how measurements were
48


made and by whom, and he discusses the disagreements among them
and why.
In the following discussion, I will describe each visual, analyze
each one using tagmemic analysis techniques, discuss the syntactical
and semantic abstraction of each graphic, and compare the relative
abstraction of each one by assigning it a position on the continua of
abstraction. The next chapter of this thesis will then explore in detail
this analysis process.
Brodie's Drawing of Coggia's Comet
Figure 8 reproduces Brodie's drawing of Coggia's Comet.
Figure 9 shows a tagmemic grid to analyze Brodie's drawing in light
of its syntax (constituent visual elements) and semantics (rhetorical
message). Figure 10 shows the graphics position on the abstraction
continua.
The bright white circle of the comet's head emits orderly arches
of white, diffuse strokes across a black background. The dominant
relationship of the simple visual elements is the contrast between
black background and white, diffuse strokes. The variations in tone
imply diffusion of material. The dynamic interplay between the
streams of white material and the black sky gives us an impression of
motion.
The text does not refer to this particular comet, Coggia's.
However, this drawing reinforces Whipple's description of cometary
halos in the first paragraph:
...Especially intriguing to me were the beautiful
"parabolic" halos shown by several comets, ... (157)
While, at first glance, this drawing is a portrait of a particular
natural feature and, thus, would be considered mimetic, it depicts the
formation of successive cometary halos.
49


Figure 8. Coggias comet of 1874 on July 13 (drawn by Brodie)
(Reprinted by permission of the Smithsonian Institution Press from The
Mystery of Comets by Fred L. Whipple. Smithsonian Institution Press,
Washington, D.C. 1985. pp. 157-167.)
Syntactically, Brodie's drawing can be placed relatively high on the
abstraction scale; the rigid arches are regularly spaced, and their only
relationship is to the comet's head and the black sky. Semantically, it
is even more abstract, dealing with the successive formations of halos.
50


Contrast Variation Distribution
Particle A bright white circle centered in the upper third of a black sky forms the comet's head. At first glance, we see a lighted Roman candle. Particle White-on-black reproduces the sight of a comet. Star fields, scales, or directional indicators could tell us where the comet is or how big it is Particle We see a description of behavior. The text discusses 19th-century observations of cometary halos
Wave The black trail from the comet's head gives a sense of motion, as do the arches of white material streaming behind the head. Wave Variations in intensity of white indicate outflowing material. Wave The arching white material flowing from the comet's head graphically depicts the idea of multiple halos successively forming.
Field The contrast of white on black and varying shades of white emanating from the bright circle are the dominant relations. Field The force of the solar wind striking the outflowing material from the comets head could be explicitly depicted. Field Arches form concentric half circles around the center of the comet while the text speaks of halos that change shape and size and that form, dissipate, and reform.
Figure 9. Analysis grid of Brodies drawing
51


Generic,
Structural,
Symbolic,
1
Levels
of
Abstraction
[syntax]
Brodie's
drawing
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 10. Abstraction continua with Brodie's drawing
52


Schmidts Drawing of Tebbutts Comet
Figure 11 reproduces Schmidts drawing of Tebbutts Comet of
1861. Figure 12 shows the analysis grid. Figure 13 presents the
abstraction continua with this drawing in position.
Figure 11. Tebbutts great comet of 1861 (drawing by J. F. Julius
Schmidt, Vienna)
(Reprinted by permission of the Smithsonian Institution Press from The
Mystery of Comets by Fred L. Whipple. Smithsonian Institution
Press, Washington, D.C. 1985. pp. 157-167.)
The same visual elements used in Brodies drawing compose this
drawing. The swirling shapes of the halos in bright tones of white
make this drawing more dynamic than Brodies drawing. The
angularity of layers of material renders this drawing more realistic
than the ordered ranks of arches in Brodies drawing.
53


Contrast Variation Distribution
Particle The comet's head is a white small dot on a black sky. Diffuse white lines of varying tones and intensity swirl around the head. A scale at the bottom provides a measure of amplitude. Particle Star fields would allow us to find the comet's location. The date of the observation could also be useful. Particle A wild succession of cometary halos swirl around the comet's head.
Wave Swirling white streaks give an impression of rotation and dissipation of material. Wave Dissipation of material seems to occur in the successive swirls. Rotation is implied by tonal variations. Wave This drawing sharply emphasizes the production of halos.
Field Swirls of white cascade out from the white dot. Variations in the white streaks show motion and dissipation. Field There is no strong sense of direction, only rotation. Field The text does not mention the figure explicitly, although Schmidt's work on other comets is discussed.
Figure 12. Analysis grid of Tebbutt's comet
The scale across the bottom could permit us to determine the
size of the comet's halos; however, the scale reproduces faintly here
and is barely decipherable on the page in Spinning Comets. The
scale seems to be an example of Schmidt's measurement technique as
described in the text:
...Meanwhile, in Vienna, J. F. Julius Schmidt, a great
double-star observer, carefully measured the diameters
of the halos across the coma perpendicular to the
direction of the tail. He made five to ten settings of his
54


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 13. Abstraction continua with Tebbutt's comet
telescopic crosswires for each measurement. From
these measures, he concluded that the halos were
growing quite rapidly and that a new one appeared every
few hours. ...(159)
55


The text mentions neither Tebbutt's Comet nor this graphic
directly; however, this drawing and the previous one, taken together,
form an illustration of a point Whipple makes in the second
paragraph:
...But why should they be perfectly round snowballs, and
exactly the same all over their surfaces? (Shades of
Aristotle?) In all likelihood, they originally grew by
accretion, from the impacts of smaller snowballs,
beginning with dust and cosmic snowflakes of some
sort. Is there any reason to think that all of them should
be alike? Of course not! Some might have been formed
in warmer regions, and might have been composed
predominantly of rocky grains. ... (157)
This particular passage, addressing 2nd stasis questions of definition,
provides a series of plausible descriptions of cometary nuclei.
Syntactically, this drawing concentrates on the surface aspects
of the comet and its swirling arches of material that vary in tone are
mimetic. Semantically, it offers information about the comet's
production of successive halos. On the abstraction continua, this
drawing is located relatively lower on the syntactic scale and higher
on the semantic scale than Brodie's drawing.
56


Struves Drawing of Donates Comet of 1858
The third drawing, in Figure 14, shows an irregular comet head
trailing a bright shroud of white material. Figure 15 is the analysis
grid of this drawing. Figure 16 shows Struves drawing on the
abstraction continua.
Figure 14. Donatis comet of 1858 (drawn by Otto Struve at Pulkovo
on October 5)
(Reprinted by permission of the Smithsonian Institution Press from The Mystery
of Comets by Fred L. Whipple. Smithsonian Institution Press, Washington,
D.C. 1985. pp. 157-167.)
57


Contrast Variation Distribution
Particle On a hazy background an irregular white comet head streams material behind it as it moves in a southwesterly direction. Particle A star field would allow us to locate the comet in the sky. Particle Only one halo is depicted.
Wave The diffusion of white gives the impression of brightness. A narrow trail behind the comet indicates forward motion. Wave Successive halos are not present. The nucleus is a white dot sporting a fan of bright material. Diffuse white circling the comet renders a bright glow. Wave The darker trail behind the comet's head suggests motion. Compass points printed on the graphic explicitly shows direction.
Field The broad streak drapes the comet's nucleus, leaving a trail to indicate motion. The diffuse circle of light represents the glow of cometary light. Field Slight variations in intensity indicate density of material in the tail, but they do not show variations in structure. Field This drawing depicts power as opposed to the other two precise structural studies. This more nearly reproduces the appearance of a comet to the human eye.
Figure 15. Analysis grid of Struve's drawing
The background is hazy. Directions are indicated by labelled
compass points in the sky. Three successively larger elements
compose the comet's structure: the tiny white dot of the nucleus, the
bright fan-shaped halo, and the white shroud streaming behind the
comet's head. Motion is implied by variations in tone in the comet tail,
giving us the impression of a trail behind the comet's head. The entire
comet glows in an oval of diffuse light.
58


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Struve's
drawing
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics] Generic,
Surface, Structural,
Mimetic Symbolic,
Figure 16. Abstraction continua with Struve's drawing
The text does speak of this comet, Donati's Comet of 1858, but
does not specifically refer to this graphic. Of the first three drawings
in the chapter, this drawing would seem to be the most portrait-like.
Only one moment in time is represented.
59


Syntactically, it uses the same elements as the other two drawings to
emphasize mimetic characteristics. Semantically, it is much less
explicitly revealing information about successive formations of halos.
Schmidts Drawing of Donates Comet
Figure 17 reproduces Schmidts drawings of Donatis Comet of
1858. Figure 18 is the analysis grid for Schmidts drawings, and
Figure 19 shows its position on the abstraction continua.
Figure 17. Donatis comet of 1858.
Left, October 6; right, October 7. (Drawn by J. F. Julius Schmidt,
Vienna.)
(Reprinted by permission of the Smithsonian Institution Press from The Mystery
of Comets by Fred L. Whipple. Smithsonian Institution Press, Washington,
D.C. 1985. pp. 157-167.)
The drawings show a pair of comet heads on a white
background. The shape of the nucleus is a textured disc with a portion
missing. The halo and tail are rendered lightly by diffuse strokes. In
60


Contrast Variation Distribution
Particle This double drawing consists of diffuse streaks of black on white. A mottled disc in the center is missing a portion. Particle Indicators of size, direction, or date could be useful. Particle The comet's nucleus carries spots and texture. The second drawing also shows structure in the halo.
Wave One day's time separate the drawings. Perhaps unintentionally, the second drawing is larger than the first. Wave Dark spots and shading indicate changes in the nucleus. Wave This drawing's details focus much more on the structure of the nucleus while the others emphasized halos.
Field Variations in diffuse streaks of black indicate the intensity of light. Small differences in the two nuclei indicate change from one day to the next. Field Black-on-white is not as dramatic as white- on-black. Size difference from day-to- day is implied, but actual size, direction, and position are not indicated at all. Field The discussion shifts from observations of halo formation to the nucleus, and this drawing focuses on the nucleus.
Figure 18. Analysis grid of Donati's comet
the drawing on the right, the shape of the nucleus remains the same,
but the texture is more pronounced. A halo forms around the
nucleus.
As Whipple elaborates his spinning snowball theory, he first
establishes that halos have been observed for several comets; then he
resolves disagreements among the interpretations of those results and
confirms that rapidly successive halos would be consistent with his
model. He then leads us to the next step which is to describe the
nucleus and what happens there.
61


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Struve's
drawing
Donati's
comet
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics] Generic,
Surface, Structural,
Mimetic Symbolic,
Figure 19. Abstraction continua with Donati's comet
The text tells how this particular comet was observed by George P.
Bond at Harvard and by J. F. Julius Schmidt in Vienna. Schmidt
concluded from his measurements that halos were forming every few
hours while other observers calculated much longer periods of
formation. The text discusses in depth Schmidt's observations and
62


their importance as a starting place for Whipple to determine the
rotation period for the comet if his theory of spinning snowballs was
correct.
The first three drawings emphasized the formation of halos.
Whipple then shifts the discussion:
Be that as it may, the clockwork precision of halo
production from Donatis comet over more than seventy-
eight periods clearly required a rotating, spinning comet
nucleus that was very active on one side and quite
inactive on the other. Therefore the actual snowball
nucleus had to be asymmetrical over its surface.
Its fast spin led me to question whether the
nucleus might fly apart because of the centrifugal force
at its equator. (160-161)
The focus of the illustrations shifts as well. The fourth drawing,
Schmidt's, concerns itself much less with successive formations of
halos and concentrates on the nucleus itself. The narrative shifts
from 1 st and 2nd stasis questions of fact and definition to 3rd stasis
questions of cause.
Syntactically, this drawing, compared to the first three, should
rank lower on the scale the contrast between dark and light is less
dramatic and more faithful to appearances; it focuses on the surface
characteristics of one part of the comet, the nucleus. Semantically, as
well, this drawing calls attention to the surface characteristics of the
nucleus. The distinctly individual disc with a missing portion
identifies this comet as Donati's in 1858. Change in the comet from
one day to another is indicated solely by differences in size and texture
of the nucleus* surface. The drawing does not represent dynamic
changes in the comet as much as it shows two separate portraits of the
comet.
63


Photographs of P/SW1
Figure 20 reproduces an outburst of P/Schwassmann-
Wachmann 1 (P/SW1) from the Oak Ridge Observatory in
Massachusetts. Figure 21 is the analysis grid, and Figure 22 shows
the abstraction continua with P/SW1 in place.
Figure 20. P/Schwassmann-Wachmann 1
Sequence showing an outburst of P/Schwassmann-Wach-mann 1, as
photographed on February 1,10,13, and 28,1981, by R. E. McCrosky
and C.-Y. Shao at Oak Ridge Observatory with the 1.5-meter reflector,
operated by a grant from NASA. Compare this outburst with that of
the same comet in 1961, shown in Chapter 20.
(Reprinted by permission of the Smithsonian Institution Press from The Mystery
of Comets by Fred L. Whipple. Smithsonian Institution Press, Washington,
D.C. 1985. pp. 157-167.)
64


Contrast Variation Distribution
Particle Four sequential photos are arranged in a block. Dates appear in the lower right comers of each Particle Some comet photos are negative images, white- on-black, to highlight structural details. Sequences can be arranged in a linearly. Particle The text specifically describes the unusual behavior of this comet.
Wave Temporal sequence is emphasized by the numbers. The explosion is the dramatic photo, but the tiny dot at the beginning and end presents scenario. Wave The beginning of a tiny, far-away comet and the end with a similar dot receding gives us a story. Wave The text describes the mysterious flaring of a barely visible comet. It clearly describes periodic explosions such as the one shown.
Field Your eye must pass over the four photos several times to establish what is the correct order. Field All four photos in a row would more clearly indicate temporal sequence. Spelling out full dates would make the temporal sequence clearer. Field The caption encourages us to go to Chapter 20 to see another explosion of this same comet. The text also promises to return to this comet.
Figure 21. Analysis grid of P/SW1
The numbers in the lower right comer of each photo refers to
the day in February 1981 when the photo was made. We see first a
small bright dot against a black sky in which only one other object can
be seen, faintly. In the second photo, a bright circle occupies the
center of a black field. In the third photo, a large, four-pointed, fuzzy,
bright form occupies center-right. Another hazy, bright object is at the
center. In the fourth photo, a small bright circle is centered and four
other oval-shaped objects float in the black sky.
65


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Struve's
drawing
Donati's
comet
P/SW1
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics] Generic,
Surface, Structural,
Mimetic Symbolic,
Figure 22. Abstraction continua with P/SW1
This illustration, a sequence of photographs taken in 1981,
accompanies the discussion of shape, gravitational forces, and
centrifugal force leading to an explanation of why comets have to be
round. As an example of the serendipitous nature of scientific
86


endeavor, Whipple employs an anecdote about the first two comets for
which he calculated the spin periods: they represented nearly the
extremes in the range of periods that he later found for comets in
general. The ensemble of the four photos show exactly what the text
discusses:
... The other was the phenomenal periodic comet
Schwassmann-Wachmann 1 P/SW1 for short
which moves just beyond Jupiter in a nearly circular
orbit. It has been a mystery since its discovery in 1927.
Although a very large comet, it is usually very faint,
observable only by large telescopes, and then not
visually. Yet, now and then, it flares hundreds of times
in brightness, becoming an object easily seen in small
telescopes. I found that its spin is quite stately, almost
exactly five days per turn, making it the longest comet
period I have measured. The average period is about
fifteen hours. (162)
This particular passage addresses the 4th stasis questions of value
the mysterious, spectacular behavior of the comet makes its study
more interesting. Addressing himself to laymen, Whipple must
explain why this comet is interesting, and he uses anecdotal
expressions and intriguing teasers:
Oddly enough, the first two comets for which I
calculated the spin periods represented nearly the
extremes in the range of periods that I later found for
comets in general... It has been a mystery since its
discovery in 1927. ...
P/SW1 presents a real puzzle, and also a key,
perhaps, to the nature of comets. ...We will return to
P/SW1 after we study other clues about the unearthly
composition of comets. (162)
67


The ensemble of photos of P/SW1 concentrates our attention on
the singular phenomenon the outburst of a particular comet. It
depicts a sequence of actual events. Its position on the abstraction
continua is very near the extreme concrete axes.
Zdenek Sekanina
Figure 23 reproduces the portrait photo of Whipples colleague,
Zdenek Sekanina. Figure 24 presents the analysis grid for this photo,
and Figure 25 shows its position on the abstraction continua.
Figure 23. Zdenek Sekanina
(Reprinted by permission of the Smithsonian Institution Press from The Mystery
of Comets by Fred L. Whipple. Smithsonian Institution Press, Washington,
D.C. 1985. pp. 157-167.)
68


Contrast Variation Distribution
Particle The photo is a black- and-white, passport- type portrait. Particle Other sorts of portrait photos could show the subject engaged in an activity. Particle This is one among several portraits of prominent researchers in the book.
Wave We see a smiling face, a static pose, and a neutral background. Wave This is a formal photo: back-lit, full-bust. Wave The smiling face relates to the text that speaks of my colleague.
Field The caption labels the photo Zdenek Sekanina. The text explains the significance of this person. Field Other photos show astronomers with telescopes or engaged in an activity. Field This one fits in with the others in an effort to personalize this research.
Figure 24. Analysis grid of Sekaninas photo
The photo is a standard, full-bust, back-lit, passport-type
photograph. It occupies the upper half of the page. Throughout the
book, one finds portraits of several other researchers: some are
portrait photos like this one, and others show subjects with
astronomical instruments. Some are portrait drawings.
This sixth illustration is a portrait photo of a well-known
researcher, Zdenek Sekanina. Whipple explains that Sekanina used
the same model at the same time as he to determine the direction of
spin. He describes their mutual contributions to the development of
the theory and their collaboration on the problems.
69


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Struve's
drawing
Donati's
comet
P/SW1
Sekanina
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 25. Abstraction continua with Sekaninas photo
The text narrates the parallel activities of Whipple and
Sekanina as they sought to calculate spin periods of two comets:
70


While I was first calculating the spin periods of
these two comets, Zdenek Sekanina, my colleague at the
Smithsonian Astrophysical Observatory, independently
applied the dirty-snowball model to the spin of comets.
He realized that, if the gas and grains from a spinning
nucleus come out mainly from the afternoon face of the
comet, the asymmetry should show up both in good
comet photographs and in visual observations. Indeed,
this is true for many comets (as I had found for comet
P/SW1). The inner coma streaks out in a direction that
gives a clue to the direction of the spin axis, the two
being more or less perpendicular to each other as seen
from the Earth. In many cases, the geometry of Sun,
Earth, and comet hides the effect. Sekanina also
realized that asymmetry in the coma could measure the
afternoon delay of the principal activity, or the lag angle
of sublimation. (162)
Whipple, here, addresses 3rd stasis questions of cause as he explains
how the asymmetry of the comets nucleus would affect the production
of halos that had been observed.
Syntactically, the photo can be placed on the concrete end of the
abstraction continua. Semantically, its purpose is to describe a unique
individual and should be considered concrete.
Graph of Enckes Comet
Figure 26 reproduces the graph of the change of period of
Enckes comet. Figure 27 is the analysis grid, and Figure 28 shows
the graph on the abstraction continua.
71


Figure 26. Graph of Enckes comet
The observed reduction in the change of period of Enckes comet since
its discovery. The curve gives the loss in period for each revolution of
3.3 years
(Reprinted by permission of the Smithsonian Institution Press from The Mystery of
Comets by Fred L. Whipple. Smithsonian Institution Press, Washington, D.C.
1985. pp.157-167.)
The axes of the graph form a box with tic marks on three sides.
The type size and weight used for legends and values are all the same.
The vertical axis measures the reduction in change of rotation period
in hours and ranges from -4 to 0. The horizontal axis marks years of
observations. The year of discovery is plainly labelled on the dark
curve of change. The sweep of the curve shows rapid change about a
century ago and a steadily declining change rate since. Against
expectations, the curve rises toward zero change since the graph plots
not a direct record of change nor the observed period length, but a
reduction of change in period.
72


Contrast Variation Distribution
Particle The axes of the graph form a box. The year of discovery is plainly labelled. Change, in hours, is reported as the observed reduction in the change of period. Particle Individual observations could be marked along the curve to emphasize the recurring nature of the observations. Particle The graph shows change over time.
Wave The individual data points are connected with a dark, thick line. Wave The dark curve contrasting with the finer lines of the axes draws the eye to the sweep of the curve. The curve rises to zero. Wave The curve shows rapid change in the past and little change recently.
Field The vertical axis extends from -4 to 0, causing the curve to rise to express less change. Field Individual data points could be shown. More precise values on the vertical axis could lend information. Field The text speaks of the rate of change as linked to the shape of the nucleus.
Figure 27. Analysis grid of Enckes graph
The text describes the idiosyncrasies of Enckes comet and
Sekaninas work to determine its spin axis. This work is presented as
a clue to determine the shape of the nucleus.
Once Sekanina had located the pole of Enckes comet
and found it so strangely placed (almost parallel to the long
axis of its orbit), we independently realized that this
peculiarity might help explain its major remaining mystery
the rapid change in its orbital period a century and a half
ago, which slowly decreased almost to zero by the 1970s.
Suppose the nucleus were not round, ... (164)
73


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Encke's
Graph
Struve's
drawing
Donati's
comet
P/SW1
Sekanina
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 28. Abstraction continua with Enckes graph
Whipple is again addressing the 3rd stasis questions of cause and
satisfying 4th stasis issues of value. The behavior of the comet is
linked to the shape of its nucleus, and the use of the terms major
74


remaining mystery provides a justification and an enticement to
continue reading.
The graph represents a fairly abstract graphic on both the
syntactic and semantic continua. Syntactically, it relies on simple
lines of varying intensity to delineate the axes and the curve. The
values are expressed in simple figures: years and hours. This
simplicity is deceptive, since the vertical axis measures the reduction
of change in negative hours. With the scale measuring negative
values, we must reconcile "rapid change" with a gentle dip in the
curve and "slowly decreased almost to zero" with the most dominant
visual feature in the graphic a steady, sharp climb of the curve.
The graph, like that of Brownlee and Kirtz in Figure 4b, shows change
over time. It should be considered moderately abstract on the syntactic
axis. On the semantic axis it is also moderately abstract.
Jet Force Diagram
Figure 29 reproduces the diagram of the effect of jet forces on the
comet. Figure 30 presents the analysis grid, and Figure 31 shows the
abstraction continua with the jet force diagram positioned on it.
The graphic is a drawing of a comet nucleus being struck by
sunlight. Labels and arrows identify comet parts and physical forces.
The texture of the irregularly shaped comet nucleus is achieved by a
combination of variations in shading and dark spots. Visually, the
primary dynamic in the drawing is the confrontation of two opposing
banks of arrows, sunlight, and ejected gas. The jet force and the
tipping force on the pole are not particularly distinguished from other
elements of the drawing. A lever arm is labelled, although its
function is not explicit nor is it mentioned in the text or caption. The
text, dealing with 3rd stasis questions of cause, introduces the
phenomenon of precession with a brief allusion to forces that disturb
the spinning of a flattened ball:
75


Pole
Figure 29. Jet force diagram
For the oblate comet nucleus in the diagram, the jet force is directed
below the center of gravity of the nucleus and so tends to tip the pole
counterclockwise, causing precession.
(Reprinted by permission of the Smithsonian Institution Press from The Mystery of
Comets by Fred L. Whipple. Smithsonian Institution Press, Washington, D.C.
1985. pp.157-167.)
i
...Suppose the nucleus were not round, but somewhat
oval shaped, a flattened ball. Internal forces and friction
in that case would make it spin about its shortest axis.
It would spin with its bulge at the equator, like a
squashed top. The jet forces that change the period
would also try to tip the spinning top or nucleus, except
when the Suns rays were shining directly on the
equator or a pole.
All spinning tops respond to a tipping force. The
result is known as precession. ... (164 and 165)
76


Contrast Variation Distribution
Particle An irregularly shaped, textured nucleus bears labels. Arrows represent sunlight striking the comet, ejected gas, and jet forces. Particle The force of the solar wind would be more apparent if ejected gas were shown streaming behind the comet. Particle Geometrical concepts, such as pole and rotation, are presented as a prelude to a discussion of precession.
Wave The seven arrows bearing down on the comet show strong motion. The arrows of ejected gas indicate direction, and the arrow around the pole shows direction. Wave The two banks of arrows confronting each other work as the main visual focus. The arrow indicating force on the pole, seems to be only a pointer. Wave Although jet forces are mentioned once in the text and prominently in the caption, they are visually subordinate to other elements in the drawing.
Field Labels identify various parts of the drawing. Field The jet force and the tipping force should be visually distinguishable from the other parts. Field There is no mention of a lever arm in the text, and jet force is mentioned without a definition.
Figure 30. Analysis grid of the jet force diagram
Syntactically, this drawing is fairly abstract; it represents a generic
comet subjected to natural forces. Semantically, I find it to be very
abstract since it should depict the dual processes of the solar wind
causing the nucleus to eject gas and the effect of jet forces from the
ejected gas on the rotation of the nucleus.
77


Generic,
Structural,
Symbolic,
Levels
of
Abstraction
[syntax]
Jet Force
Struve's
drawing
Brodie's
drawing
Donati's
comet
Tebbutt's
comet
P/SW1
Sekanina
Individual,
Surface,
Mimetic
object area quantity process
[semantics]
Generic,
Structural,
Symbolic,
Figure 31. Abstraction continua with the jet force diagram
78


Wobble Cartoon
Figure 32 reproduces the pair of drawings describing the motion
of a comet or other body in space. Figure 33 is the analysis grid of the
cartoon, and Figure 34 shows its position on the abstraction grid.
Figure 32. The wobble of precession
Left, as the top tries to fall over, its pole wobbles in the same sense as
its rotation; right, as the Moon and Sun try to tip the pole
perpendicular to its orbital plane, the pole precesses in the opposite
sense.
(Reprinted by permission of the Smithsonian Institution Press from The Mystery of
Comets by Fred L. Whipple. Smithsonian Institution Press, Washington, D.C.
1985. pp.157-167.)
This drawing extends the metaphor of a top. The two drawings
are executed in fine black lines, straight and curving arrows, and
labels. In the simple case on the left, the pole of a spinning top wobbles
in the same sense as its rotation. The drawing on the left shows a
Pole of Earth's OrDit
79


Contrast Variation Distribution
Particle Two line drawings represent simple and complex cases of the wobble of precession. Particle The parallels between the top and Earth could be more visually explicit, perhaps by superimposing a top on the Earth in the second drawing. Particle Motion is conveyed by the arrows. The dotted line of Earth orbit suggests a direction
Wave The relatively simple top on the left follows the text as a prelude to the more complex case illustrated on the right. Wave Curving arrows show rotational motion. Arrows on the right attract attention to gravitation forces. Wave The circling arrows around the dotted lines of the poles express the analogy visually.
Field The complexity increases from left to right. Field The Earth on the right could more closely emulate the top, or the poles in both drawings could be more nearly similar. Field The text makes explicit the metaphor: ... Earth is spinning like a top,...
Figure 33. Analysis grid of the wobble cartoon
simple top with two bands of decoration. Irregular, more free lines
indicate shadows under the top. Dashed lines represent its pole and
axis. Curved arrows show motion around the center of the top. A
circle with arrows indicates how the pole moves. A more complex
scenario diagrammed on the right indicates the effects of gravitational
forces on the orbit of Earth. The drawing on the right is more complex
with ovals, arrows, lines and labels.
80


Generic,
Structural,
Symbolic,
Levels
of
Abstraction
[syntax]
A
Wobble [R.]
Enckes
Graph
Jet Force
Wobble [L.]
Struve's
drawing
Donati's
comet
P/SW1
Sekanina
Brodie's
drawing
Tebbutt's
comet
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 34. Abstraction continua with the wobble cartoon
The metaphor of the top is explicit:
All spinning tops respond to a tipping force. The
result is known as precession. Gravity makes a top fall
over when it is not spinning fast enough. Once set
81


spinning, the top keeps wobbling around, its spin axis
turning around the vertical. The Earth is a familiar
astronomical example of a disturbed gyroscope, or
spinning top. ... Because the Earth is spinning like a
top,... (165 and 166)
Whipple again concerns himself with an audience of lay people
who may or may not be familiar with the scientific method as he
describes how he and his colleague solved the problem:
Sekanina and I joined forces to see whether we could
apply these ideas to comet Encke and make sense out of
its strange motions. We coded our computers to
calculate the jet forces and the motion of the pole for an
oval nucleus that is relatively dark and inactive over its
southern hemisphere. We found that we needed to know
the spin period, which I found to be about 6'/2 hours. To
our great satisfaction, our calculations fit very well with
the observed variable changes in the period. The
nucleus came out slightly flattened, a small percentage
of its radius at the poles. The calculated amount
depended upon the diameter we assumed, which we took
to be about 3 kilometers. If the actual diameter is
greater than this, the flattening will increase
proportionately. (166)
Celestial Map
Figure 35 reproduces the celestial map showing Sekanina and
Whipples predictions for Comet Encke. Figure 36 is the analysis grid
of the graphic, and Figure 37 shows its position on the abstraction
continua.
A gridded oval map with latitudes and longitudes indicated
represents the sky. The path of the precession of the pole of Comet
Encke through time follows a curve with points labelled from 1000 BC
to 3000 AD.
82


0#
Figure 35. Celestial map
Motion of spin of pole of Enckes comet across the sky. Solid curve
represents period of observations. (By Fred L. Whipple and Zdenek
Sekanina.)
(Reprinted by permission of the Smithsonian Institution Press from The Mystery of
Comets by Fred L. Whipple. Smithsonian Institution Press, Washington, D.C.
1985. pp.157-167.)
Having described a cometary nucleus that would fit the
observations, Whipple can now explain the major mystery of why the
comets behavior:
83


Contrast Variation Distribution
Particle An oval, gridded map of the heavens shows celestial coordinates. The solid curve marks recorded observations of Enckes Comet, and the broken curve represents postulated positions. Particle Some celestial maps show constellations or positions of planets. Since Encke is unusual perhaps a comparison with and average comet would be in order. Particle The text speaks of large motion.
Wave The hooking curve across the bulging grid leads the eye across the map and to the pole. Wave Large motion is almost a degree per year over the nearly two centuries since Enckes comet was discovered. Wave The curve shows change of position more clearly than rate of change.
Field The modem era is shown in solid lines; speculative periods are in broken lines. Field years seem most important because rate and amount of change must be determined by measuring. Field The explicit rate of change mentioned in the text is not readily apparent from the illustration.
Figure 36. Analysis grid of the celestial map
... To our surprise, the pole apparently had
become almost stuck in one direction for several
hundreds of years before 1700. ... Since the 1700s, the
south pole has turned around and become well radiated,
but not long enough for the insulating layer to be cleared
off. ...(166 and 167)
He is addressing 3rd stasis questions of cause, 4th stasis questions of
value, and, finally, opens the question of policy in the 5th stasis, as he
concludes:
84


Generic,
Structural,
Symbolic,
A
Levels
of
Abstraction
[syntax]
Wobble [R.]
Map
Encke's
Graph
Jet Force
Wobble [L.]
Struve's
drawing
Donati's
comet
P/SW1
Brodie's
drawing
Tebbutt's
comet
Sekanina
object area quantity process
Individual, [semantics]
Surface,
Mimetic
Generic,
Structural,
Symbolic,
Figure 37. Abstraction continua with the celestial map
... As an alternative to our theory, of course, the
southern hemisphere may be naturally rocky, because of
the core of Enckes comet may have been formed that
way. The question is a vital one, as we shall see later,
because large, old, tame comets may possibly turn into
asteroids if their cores are truly rocky. Will we ever
know for sure? A space mission to the comet could
answer the question. (167)
85


Summary
We have now assembled the components for a detailed
description of Fred L. Whipples Spinning Comets. We can describe
the physical entity that is the book, The Mystery of Comets. We can
discuss the rhetorical strategies chosen to address a lay audience. We
can also describe the graphics as rhetorical elements of the text. We
have examined the graphics in terms of their relative abstraction and
have compared them with each other. The next chapter discusses this
analysis and uses all of this information to describe how the graphics
interact with the text, that is, how they fulfill their role as rhetorical
elements.
86


5. DISCUSSION
In front of a Cubist work of art, the spectator was to
realize that no single interpretation of the fluctuating
shapes, textures, spaces, and objects could be complete
in itself. And, in expressing this awareness of the
paradoxical nature of reality and the need for describing
it in multiple and even contradictory ways, Cubism
offered a visual equivalent of a fundamental aspect of
twentieth-century experience.
Robert Rosenblum
The design and effective use of graphics involve complex,
interrelated problems that admit very few simple solutions. While I
have not been able to discern overarching rules about the design of
graphics in all cases, analyses such as those in the previous section of
this thesis should prove helpful in guiding the use of graphics in
different rhetorical situations. As designers, we can use the concept
of abstraction to approach those two key aspects of any communication
problem: what to say and how to say it.
In the previous chapter, stasis theory and tagmemic analysis
techniques yielded a large quantity of information about particular
graphics, their relative semantic and syntactical abstraction, and
their rhetorical context. Each graphic of Whipples Spinning
Comets has been studied in its unique context, and that context
influences where the graphic appears on the abstraction continua.
The placement of a graphic along the continua, of course, is entirely
subjective, and each graphic is more or less abstract relative to
another.
Let us now summarize these analyses. First, we will describe
the rhetorical context as we now know it. Given the information from


the mechanical and rhetorical analyses, we can assess how Whipple
presents his message, how he organizes his text, and the strategies he
uses to reach his audience. Then, we will review the salient traits we
have discovered in the set of graphics in Spinning Comets. We can
now discuss the role that each graphic plays as a rhetorical element.
Finally, we will also explore some of the questions this work has
raised.
The Rhetorical Context
ParticleWhat It Is
The information we have from the analyses of the previous
chapter allows us to describe the rhetorical context of Spinning
Comets. The Mystery of Comets is a small, inexpensive paperback
copiously illustrated with black-and-white photos, drawings, and
diagrams. A center section carries color photos. The chapters of the
book first review pre-20th-century knowledge about comets, then
recount the development of modern theory and observations, and
speculate on the benefits of further studies. The ten pages and ten
illustrations of Spinning Comets tell how Whipple and his
colleagues developed their descriptions of cometary nuclei and the
processes that occur on them.
WaveWhat It Does
Using information from the rhetorical and mechancial
analyses, we can describe Spinning Comets as a dynamic entity.
Spinning Comets requires a moderate amount of effort to read and to
88


understand. According to the mechanical analyses that are
summarized in Table 4, it should be accessible to its intended audience
of lay people with a minimum of a high-school-level education.
Features of the text that accommodate non-scientific readers
include
a mix of long and short sentences (about 20% long sentences and
an equal percentage of short sentences),
the predominance of active voice (about 87% of the sentences are
active), and
an effort to personalize the material.
The various readability formulas computed by the software are:
Flesch Reading Ease 53
Gunnings Fog Index 14
Flesch-Kincaid Grade Level 11
These formulas are based on measureable parts of the text:
number of words and sentences, length of words and sentences,
passive constructions, and so forth. Klare (1974) has stressed that
readability formulas provide good indices of difficulty but that they do
not indicate causes of difficulty or show how to write readably. In the
case of Spinning Comets, these scores indicate that an interested,
fairly well-educated, lay audience will be able to read the text without
enormous trouble.
FieldHow It Works
The restrained use of the passive voice, 16 passive sentences out
of 121 total sentences, enhances the accessibility of the text. Research
shows (Felker 1980) that in most common contexts active sentences are
more readily understood and recalled. Whipple also uses direct
questions in six sentences and and one exclamatory sentence. In
89