Psychophysics and the visual perception of geotechnical illustration

Material Information

Psychophysics and the visual perception of geotechnical illustration
Carter, Lorna M
Publication Date:
Physical Description:
xi, 184 leaves : illustrations, maps (including 1 color photograph) ; 29 cm


Subjects / Keywords:
Communication of technical information ( lcsh )
Visual perception ( lcsh )
Communication of technical information ( fast )
Visual perception ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 175-180).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science in Technical Communication, Department of English.
Statement of Responsibility:
by Lorna M. Carter.

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:
22692988 ( OCLC )
LD1190.L67 1989m .C37 ( lcc )

Full Text
Lorna M. Carter
B.A., University of Nebraska, 1963
M.A., University of Nebraska, 1965
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 in Technical Communication
Department of English

1989 by Loma M. Carter
All Rights Reserved

This Thesis for the
Master of Science in
Technical Communication degree by
Lorna M. Carter
has been approved for the
Department of
Charles E. Beck
Martin Tessmer
Date ///£?/?/

I. INTRODUCTION..................................... 1
Background of the Study........................ 2
Development of the Problem..................... 4
Direction and Limits of the Discussion......... 6
GEOLOGIC ILLUSTRATING......................... 12
Dots....................................... 16
How People React to Dots in
Psychophysical Experiments............... 16
Dots as Important Elements of
Geotechnical Illustration............. 19
Lines...................................... 2 2
How People React to and Perceive Lines
in Psychophysical Experiments............ 22
Lines as Important Elements of
Geotechnical Illustration............. 26
Patterns................................... 30
How People React to Patterns in
Psychophysical Experiments............ 30
Patterns as Important Elements of
Geotechnical Illustration............. 36
Symbols and TypeRecognition and
Positioning................................. 43
How People React to Symbols and Type
Letters Used in Psychophysical
Experiments.............................. 44

How Geotechnical Illustrations Use Symbols
and Type.................................. 53
Color........................................ 58
How People React to Color, from
Psychophysical Experiments.............. 58
Color as an Important Element of
Geotechnical Illustration................. 62
Illusions...................................... 67
How Psychophysical Test Subjects React to
and Perceive Deliberate or Accidental
Illusions in Visual Material................. 67
The Presence of Illusion in Geotechnical
Illustration............................ 7 4
COHERENCE...................................... 84
How Psychophysical Research Underlies
Effective Layering in Geotechnical
Illustrations.................................. 86
Producing Clear, Coherent Layering in
Geotechnical Illustration.................. 91
Visual Elements as Used in Geotechnical
Illustration................................ 104
Dots....................................... 104
Lines....................................... 107
Patterns.................................... 121
Symbols and Type............................ 129
Color....................................... 138
Illusions................................... 148

Achieving Clarity and Coherence in Layered
Geotechnical Illustrations............ 154
FURTHER WORK............................ 167
BIBLIOGRAPHY................................... 176
A GRAPHIC EXAMPLES........................ 181

1. Summation of findings of psychophysical research
on visual perception of graphic elements...........

1. Dots used as shades to indicate three dimensions.. 106
2. Use of dots as "screening," to place section lines,
roads, creeks into the background of illustration.... 108
3. Use of dots as "screening," to place section lines,
roads, creeks into the background of illustration.... 109
4a. Use of lines to locate features and divide units..... 112
4b. Redrafted version of Figure 4a with more
differentiable patterns................................. 113
5. Effective use of lines to frame type and make type
easy to locate and read................................. 115
6. Effective use of lines to frame grouped dots............ 116
7. Ineffective and distracting use of a line grid....... 118
8. Example of too many curved lines and too much
screening............................................... 120
9. Example of well-selected, clearly-differentiable
patterns................................................ 122
10. Bold black patterns and dark lines that rise out of
the illustration's surface and create an illusion.... 125
11. Example of lines placed too close together to resolve
one-by-one, resulting in a pseudo-pattern............... 127
12. Example of patterns that mimic the makeup of
rock units.............................................. 128
13. Use of shade pattern to evoke spread of volcanic
ash cloud............................................... 130
14. Use of symbols and symbol groups........................ 132

15. Example of letters and numerals used to locate
points on a graph................................. 134
16. Example of masking and occluding of symbols......... 136
17. Use of color to evoke memory biases................. 140
18a. Distracting use of color and pattern............... 142
18b. Same area as figure 18a; redrafted to correct color
assimilation and pattern duplication problems..... 143
19. Color photograph where color printing is
ineffective......................................... 147
20a. Illusory "chicken-wire" grating across illustration
that blocks processing of information............... 150
20b. Redrafted version of Figure 20a...................... 151
21. Example of an illusory pattern and of several
shade patterns that are not clearly differentiated... 153
22. Example of multilayered illustration in which
layers cannot be easily distinguished or
processed........................................... 156
23. Example (with Figures 24 and 25) of an attempt
at "horizontal layering"--distributing complex
information into several illustrations.............. 159
24. Same over-all geographic area as Figure 23 with
dense geologic information in one small area...... 160
25. Same geographic area as Figure 23 with dense
geologic information in a second small area....... 161

Carter, Lorna M. (M.S.T.C., English)
Psychophysics and the Visual Perception of Geotechnical
Thesis directed by Assistant Professor Colleen E. Donnelly
Research in human visual perception uses visual
elements that are the same as those which appear in
scientific and technical illustrations. Dots, lines, symbols,
grids, boundaries, arrangements of type letters, and color are
visual elements that experimenters have used to test human
perception; singly and in combination, these elements also
present the information of graphic technical communication,
here, illustrations in geology. What researchers have
postulated about visual perception is useful for improving the
effectiveness of such illustrations for their audiences.
Results of psychophysical experiments, including
perception problems subjects experienced with test
materials, suggest how authors, editors, draftors, or readers
prepare and attempt to understand individual as well as
combinations of geotechnical illustrations. Illustrations from
geologic reports, containing layers of detail composed of
visual elements identical to those used in psychophysical
research, reveal problems and successes in processing and
integrating information analogous to those reported in
psychophysical testing. That so much variation in response

occurs in testing suggests that some geotechnical illustrations
communicate more effectively than others because visual
elements are used in ways that are more easily perceived.
The processing success or problems common to testing and
geotechnical illustration also suggest that standardized
drafting practices intuitively reflect principles of perception,
that drafting practices become standard after such practices
are seen to be effective, and that they are effective because
people speedily process the information they present.
Integrating results of numerous experiments thus suggests
ways to decrease confusion and speed understanding for
audiences of geotechnical illustration.
The conclusion of this thesis is that preparers of
graphic material can learn from psychophysical research to
better understand both authors' and audience's needs as well
as to apply existing drafting standards with more continuity
and consistency. Among broader issues remaining is a more
systematic relation of perception, symbology, and technical
illustration: specifically, how symbols in complex illustrations
elicit processing in graphic material and how they lead to
capturing of meaning.
The form and content of this abstract are approved,
recommend its publication,
Colleen E

Research on human visual perception, such as reported
in the journal Perception & Psychophysics, tests perception
of visual elements such as dots, other variously oriented
symbols and type letters, lines, particular color values, and
colors. Many experimenters asked subjects to look at these
materials to elicit a specific direct reaction, and thus they
were not eliciting a reaction of hesitation or perception block
because of confusion. In other experiments, researchers
employed visual illusions (produced by confusing symbols,
confusing combinations, or moving elements) to examine
potential perception problems. All the experiments
employed these visual materials without context: the visual
elements did not appear as part of a coherent design to
convey meaningful information, but were used singly, to test
reaction to the elements alone.
These visual materials, used in experiments without
context, correspond to the visual materials that authors,
editors, and draftors use every day IN context, to
communicate with readers of geotechnical illustration--
illustrations about geology. This thesis investigates and

assesses the relationship between visual perception in
psychophysical experiment and perception of the same
elements in graphic illustration, in particular demonstrating
the usefulness of perception research in preparing effective
geotechnical illustrations.
Background of the Study
Editors and "draftors" of geotechnical illustrations have
a long-standing tradition and strong convictions about when
an illustration is clear and when it is garbled and confused.
These ideas are based on long experience with "how
illustrations should look" (that is, what kinds of information
geology presents and how they combine) and on experience
with often long-established standards in the workplace. Many
editors and draftors also sense the actual audience for a
report's illustrations: where its members work, what their
level of understanding is, and how and in what chunks they
receive and process information. Authors, on the other hand,
when preparing rough illustrations that will later be
professionally drafted, occasionally seem more interested in
clumping their data together as densely and in as few
illustrations as possible. They are sometimes more
concerned with getting their illustrations in some form to
satisfy a deadline, and less concerned with any audience's
needs. Such authors may select ways to represent graphic

data that are clear to themselves, because they are so familiar
with their work, but that confuse or alienate those who seek
information in a finished report.
Although deadlines and overfamiliarity with their
material may be unavoidable for authors of scientific
illustration, the primary goals of scientific illustration remain
clarity (that the reader can sort the material) and
effectiveness (that the reader can gain the message). When
professional draftors take authors' copy and prepare an
illustration, they use drafting standards that aim for clarity
and effectiveness in the finished work, as well as for internal
consistency within an illustration and throughout a series of
similar illustrations. The goals of clarity, effectiveness, and
consistency may be more attainable, and geotechnical
illustrations may communicate better, if all illustration
preparers, from author through draftor, examine five key
issues: (1) how authors perceive and communicate their
information as they prepare an illustration or set of related
illustrations: (2) what editors and draftors do with what
authors give them, and how they know to do it; (3) what
audiences need and what they are capable of processing: (4)
what audiences have taught themselves to look for; and (5)
current technical standards for illustration. One tool for such
an integrative examination of the many components

impacting on reading and processing illustrations is
psychophysical experiments in human perception.
Development of the Problem
As a long-time geologic editor, I have recognized that
some illustrations of geologic information carry and integrate
a great deal of data clearly and effectively; others seem
cluttered but do convey their meaning after more effort on
the reader's part. Still others fail no matter how much work
is lavished on them. They remain turgid, or vague, locking
their meaning within their graphic depths. The experiments
in Perception & Psychophysics, using visual elements
analogous to those used in geotechnical illustrating, provided
information about what constitutes effective illustrating. The
argument of this thesis is that visual materials used in
perception experiments without context can be extrapolated
and given context in geotechnical illustrations. The
experimental implications are thus recognizable within
specific illustrations from published reports. These
experiments, coupled with my experience with geotechnical
illustration, suggest why people struggle to process some
illustrations but quickly amass information from others, and
how authors, editors, and draftors can consciously improve
illustration clarity and effectiveness.

The matter of improving clarity is a practical one.
Scientific illustrating is becoming more complex, as more
kinds of available data must be integrated and presented to
the public. For example, geologists in the field now map rock
units and structures, and gather other data more often by
helicopter than on foot or horseback; thus more data is
available faster. Satellite images are now available for
interpretation; improved well-logging instruments provide
new data about the interior of the earth; more sophisticated
hardware for earthquake recording and interpretation gather
information and exchange it worldwide; computers aid in
data manipulation, organization, interpretation, and transfer.
More sophisticated drafting techniques, such as graphics
computer packages, are available to speed visual information
to audiences. Because producers of technical information
seem to have more information coming in to prepare, and
because the information seems increasingly varied and
sophisticated, it seems useful to constantly examine
processes by which such information will appear as printed
illustrations. One way to make visual information
communicate better is to investigate how people take in,
process, organize, store, and recall visual information.

Direction and Limits of the Discussion
The research base of this thesis is experiments in the
process of visual perception, the process by which people's
visual systems accept photons of light, transduce them into
electrochemical signals, transmit these to the brain,
construct percepts, integrate them, "see," and understand
what they are "seeing." In brief, photon waves, or
wavelengths of electromagnetic energy, enter the pupil of the
human eye and are focused by the lens onto the retina, the
rear surface of the eye. Photoreceptor cells (including the
familiar "rods" and "cones") in the retina contain substances
called visual pigments, which absorb incoming photons,
translate them photochemically into neural responses, pass
those responses on to the next visual stage, and reconstitute
themselves in order to catch the next available photons. The
neural responses from the photoreceptors are gathered by
neural ganglion cells, which pass the responses on to the
optic nerve and thereby to the brain. Gathering, combining,
and refining of the responses takes place constantly, from the
first photochemical transferrence henceforth. Neurons at
refining points along the optic tract (the term for the optic
pathway after the optic nerves meet and cross on their way to
the brain) are increasingly selective in their response; that is,
they respond to impulses from specific parts of the earlier

response system. For example, neurons of the lateral
geniculate nucleus of the thalamus, an intermediate
processing area between the eye and the brain, fire in
response to (and transfer) impulses only from photoreceptor
and impulse-gatherer cell groups located at specific points in
the retina. The impulses eventually reach the occipital lobe
in the rear of the brain, and are distributed into areas of the
visual cortex, ultimately for final processing of patterns,
colors, and the distinguishing of percepts and objects. In the
visual cortex, "a rather direct topological (point for point)
mapping of the external visual world exists, with specific
points in the environment corresponding to specific points in
the cortex" (Coren, Porac, and Ward, 1984, p. 80); thus, areas
of the brain and areas of earlier points in the processing
system are closely linked. The "whole brain" does not
respond to an image from one area of the visual environment,
but rather very specific systematic parts make the response.
That the gathering, combining, and refining of information
continues constantly throughout the system, and that the
brain cortex culminates the refining so specifically shows that
the brain is a very systematic, carefully organized, and
delicate mechanism, containing duplication and back-up
systems, for encoding and decoding, arranging, distributing,
storing, and retrieving visual information. And it is a part of
this extremely complex process, the perception of particular

visual elements used both in research and in geotechnical
illustration, that my thesis comments on, to show how, from
experiments in perception, familiar visual elements
commonly used in illustrating geology may communicate
effectively or may fail to communicate, depending on how
they are prepared and combined.
The direction of this discussion is from details to
synthesis: I first describe readings from psychophysical
research by type of visual material used in the experiments,
and describe the relationships of the experimental elements
compared to elements used in geotechnical illustration. In
the next section, I analyze the success or failure of specific
geotechnical illustrations using criteria suggested by the
experience of researchers and their subjects. Here I begin to
generalize some principles about illustration perception and
its problems, and begin to concentrate on the subject of
information layering, a vital technique of geotechnical
illustrating. Information layering technique influences
perception of illustrations: layering can speed or slow
processing, clarify or cloud understanding. This section on
layering thus bears on drafting standards and procedures and
how to increase illustration effectiveness. Finally, I
summarize my journal observations and illustration analysis,
and make some recommendations both for effectively

preparing and integrating layers of geotechnical information
in terms of psychophysics, and for future work on the subject.
A few explanatory remarks on the possible limitations of
this study should be made here. First, psychophysical
experiments such as reported in Perception & Psychophysics
represent and test short-term perception; in most
experiments, subjects see images for milliseconds, in
controlled settings. In a real-world setting, one can reread a
figure or re-run a video, and can choose or enhance her
surroundings. Thus she can control and reinforce her own
processing. However, a reader of a technical report inevitably
receives first, "instantaneous" impressions of the illustrations-
-singly and as a groupwhich influence her toward fast and
eager processing or toward slowed and frustrated processing
of the graphic information. Thus, readers have a very rapid
reaction to illustration material in context, analogous to the
instantaneous reaction provoked in experimental situations
(even though processing within a context is influenced more
by top-down processing, and processing without context is
more sensory, influenced by bottom-up processing).
Materials within a context also produce expectationsmuch
more so than elements presented in isolation in an
experimental setting. Experienced readers of geotechnical
illustrations can have high expectations about what kinds of
information will appear and how the resulting illustration will

look. Meeting these expectations may speed processing
greatly, but processing may be impeded if the reader expects
something that is NOT present. That the experimenters'
aims and mine differ is also a reason for not being concerned
about the time element. What the experimenters learned
they related to perception as a process but not to perception
of contextual material or expectations of it, concerns which
are paramount here. Both because "instantaneous" reactions
occur and because expectations can decrease reaction time, I
have considered the length of presentation time between
experimental setting and reading of scientific illustration as
only a minor detriment to the major findings of this study.
Another possible criticism of my study is that reading
and building theory apparently from only one journal may
seem one-dimensional. This is a valid criticism. However, in
this thesis I attempt not to exhaust the topic of technical
illustration and psychophysics, but to begin to establish the
relations between them. Other resources, including
professional experience and research in linguistics and
rhetoric that deals with technical communication, have given
me considerable background in how people process
information. Combining my professional knowledge of
illustration and my prior research with research from a
journal dedicated to publishing psychophysics has placed me
squarely where I want to be to begin my examination.

Finally, I am aware of using "we" and "people" in
discussing (1) experimental subjects noted in the literature;
(2) authors, draftors, and editors of illustrations; and (3)
readers of geologic reports containing such illustrations. In
this "we" and these "people," I assume a considerable degree
of literacy and of familiarity with written word, alphanumeric
characters, a variety of standard scientific symbols, and visual
information itself. These "people" are probably Americans or
Western Europeans familiar with American culture and
science; these groups do not represent a universal humanity.

Certain visual elements that researchers used in testing
aspects of human perception were identical to the materials
that scientific authors and illustrators use or encounter in
preparing graphic representations of geologic information.
This chapter does the following things: (1) presents a
summary table (Table 1) of the findings of the experimental
research: (2) identifies and describes the visual elements
used in experiments--dots, lines, patterns, symbols and type,
color, and problem-causing illusions; (3) discusses how
psychophysical researchers have examined these elements in
experiments and how subjects perceived them; and (4) notes
how experimental subjects' perceptual problems and
successes relate to ways those elements succeed and fail in
geotechnical illustration.
The visual elements often overlap in the ways they
function: for example, dots function both as symbols by
themselves and as the constituents of other symbols and of
lines. Lines appear solid but may be composed of dots; such
lines may in some cases be perceived as solid and in other
cases as sequences of distinguishable dots. In a similar way,

patterns can consist of dots arranged in recognizable
geometric sequences repeating themselves across two-
dimensional space, or patterns can consist of other repeated
symbols, which themselves may be made of dots. Pigmented
colors, when viewed with a lens, consist of dots--in pigment,
combinations of magenta, cyan, yellow, and black. Difficult-
to-resolve line segments, symbol clusters, or dot groups may
present the viewer with illusions, perceptual problems in
sorting or integrating visual material. Thus, because dots
seem to be the archetypal element and the underlying fabric
of the visual materials, the following discussions also
interweave. For example, research articles reported in the
section "Dots" reappear in other sections on lines, patterns,
type, colors, and illusions, because they address what seems
the most generalized visual material of geotechnical
illustration, and because they apply to the way the more
combinative visual elements appear in and influence
geotechnical illustration.

Table 1. Summation of findings of psychophysical research
on visual perception of graphic elements
Dots: Dots are elementary, basic visual elements that people
perceive easily and readily. Researchers testing how
people recognize series of dots found that how quickly
subjects successfully performed perception tasks was
influenced by how close the dots were, how regular they
were, and in a formed pattern how symmetrical they were.
Lines: Lines present more definite, defined outlines than
dots do, and subjects could be tested on how quickly they
perceived, completed, and followed the direction of those
outlines. People construct more completed and complex
percepts from lines than from dots, and lines attract more
of people's organizing and completing capacity than do
Patterns: Patterns require perception of a complex,
physically larger visual area, rather than of an outline.
Some pattern combinations are quickly perceived; others
are distracting or illusory. People readily differentiate
patterns from a background but react differently to
different intensities of pattern versus background. Some
combinations assimilate and are difficult to process: others
present such stark, perceptually-distracting contrast that
the reader cannot integrate them.

Table 1 (Continued)
Symbols and type letters: Symbols (such as triangles,
squares, plus signs) and type letters convey meaning or
provide information, and people readily give their
attention and interest to them. Reaction time to locate
and recognize letters and symbols was affected by
placement and angle of a test symbol or letter; size,
thickness, and intensity of the element; and spacing of
elements in the visual field. Spacing may be especially
important. Processing of complex symbol or letter groups
is probably both from whole to part and part to whole.
Color: Color perception is an integral part of how people
relate to their environment. People develop preferred and
remembered colors for items; color perception is not
constant but changes with changes in people's physical
system. Adjacent colors in a display can assimilate, that is,
be seen as looking like each other; and spacing between
color items in a display influences reaction time for
locating and recognizing.
Illusions: Illusions in visual material distract the viewer and
cause reaction time for locating, recognizing, and
integrating visual material to slow. Completing fragmented
visual stimuli is useful and effective in people's daily lives
but can result in misleading or confused percepts. Illusory
movement, illusory boundaries, and distractor elements
that mask target elements produce further illusion
problems in visual materials.

How People React to Dots in Psychophysical Experiments
Numerous psychophysical experimenters have studied
the perception of dots, using dots in many forms: solitary
dots of various sizes, moving dots, dots in combination with
other symbols, and lines composed of dots, which the subject
may perceive as solid lines, as gray lines (dots which do not
run together but have a slight amount of space between
them), or as dotted lines (space between dots is enough so
that individual dots are perceived). The dot experiments had
the following aspects in common: (1) researchers believed
that dots are elementary and basic visual elements that
people perceive easily and readily. (2) Researchers wanted to
test how people recognize patterns created by a series of
dots, and in particular what perception mechanisms were
responsible for the reactions expected or observed. (3)
Researchers found that how quickly subjects successfully
performed perception tasks was influenced by how close the
dots were, how regular they were, and in a formed pattern
how symmetrical they were.
One research group, recognizing that "the visual system
[of an individual], when presented with a set of discrete dots,
tend[s] to organize into contours those dots that are adjacent,
similar, and (roughly) collinear," experimented to see if

people recognize patterns by using many criteria
simultaneously, or by successive brain operations that widen
their understanding of a dot array from perhaps an initial two
dots (Zucker, Stevens, and Sander, 1983, p. 515). Zucker,
Stevens, and Sander questioned whether recognition factors
for perceiving dots work simultaneously or in a sequence, and
stated that contours are perceived by the retina, not globally
but locally; further processing is needed to perceive a
continuous contour from a series of dots.
In another experiment, Jenkins (1983) stated: "If a
random-dot texture is reflected about a given axis, the
resulting bilateral symmetry is immediately detected" (p.
433). Jenkins also noted that the alignment along an axis of
symmetry seems to affect perception. In some experiments a
vertical axis allowed faster recognition than a horizontal one,
whereas the situation reversed in other experiments. In
either case recognition of an axis of symmetry seems to aid
perception. Jenkins (1983) reasoned that recognizing
bilateral symmetry in a dot pattern may work because the
visual cortex may have an organization of "columns of like
sensitive cells" (p. 433) and receptive fields in these cells,
but he noted that as yet "no direct neurophysiological
evidence" exists for "symmetrical neural organization in the
mammalian visual system" (p. 433) that would be set up for
automatic electrical-impulse recognition of symmetry. If

there is no "symmetrical neuron organization centered about
the fovea" (p. 439), then perhaps dot-pattern recognition
works through
first, a process that detects the orientational
uniformity of the component point-pairs,
irrespective of their size; second, a process that
fuses the most salient point-pairs into a salient
feature; and third, a process that determines
whether this feature is symmetric (p. 439).
People "reconstruct percepts" from sparse stimuli, such
as groups of dots or dots spaced to simulate line segments:
we can very definitely "fill in, close, and perceptually
complete partial stimuli" (Uttal and others, 1988, p. 223).
Uttal and others quoted research that
showed that observers viewing stimuli that were
close approximations of alphabetic characters
perceptually created stimulus components that
were not physically present, but were only
suggested by the arrangement of the other parts
of the stimulus (p. 223).
Furthermore, "The ability to perceive wholeness in a world of
naturally masked stimuli and noisy or incomplete views of
objects is obviously a very valuable skill" (p. 223).
People also use their capacity for depth perception and
peripheral vision to perceive and process dots: in one
experiment subjects were to use straight-ahead and also side
vision to adjust lighted lines of dots so that they were parallel
and then so that "each lateral pair of lights appears to be

1 9
separated by the same distance" (Indow and Watanabe, 1984,
p. 149).
From the studies reported, researchers believe that
dots tell them something useful about visual perception, that
people very actively and quickly perceive and recognize dots
singly and in textures, when those dots are stationary as well
as when they are moving. And dots are as important and as
basic an element in geotechnical illustration as they seem to
be in psychophysical experiments.
Dots as Important Elements of Geotechnical Illustration
Dots are very common in geotechnical illustrations.
They locate points that readers are intended to find, get
information from, and correlate on maps. For example, dots
indicate town locations used for geographic references, and
localities of a specific concentration of a chemical element
(like tungsten or molybdenum). Dots also appear commonly
on graphs to indicate the position of a particular sample with
respect to the parameters of the X and Y axes.
Dots also result from screening graphic material, a very
common procedure when illustrations layer several types of
information. (See Appendix A for an example of solid and
screened visual material. The county lines, city outlines, and
names are screened, whereas the sample locality symbols and
numbers are solid black.) Topographic contours are normally

screened in a complex map illustration that layers line
information, symbol information, and black-and-white or
color patterned information. The draftor draws the
topographic contours of a map area on a separate piece of
material, and when a negative is made, a piece of finely
dotted material is added over the contours so that all the
lines on the final negative consist of gray rather than solid
lines. The gray lines resolve into dots when examined with a
The process of screening is also necessary for printing
photographs in books. Whereas photographs from a camera
are exposed and developed in a continuous range of tones,
produced by varying amounts of metallic silver in the
emulsion process, printing can only be done through
reducing the photograph to a complex of inked dots. A
screen is placed over the photograph, an image is made from
that composite, and that dotted image (called a halftone) can
be inked in the printing process. Depending on the quality
of the photograph, the paper to be used in printing, and the
resolution needed, technical photographs are commonly
screened at 150, 200, or 300 lines per inch; and an image
made of varying intensities of dots rather than a continuous-
tone image results. (See Appendix A for examples of
screened halftone photographs.) Dot perception may
influence the success of processing in situations where one or

another screen produces a clearer or a vaguer image. A
sparser screen (such as 133 lines per inch, which printers
sometimes use instead of 150) may not allow the reader to
complete, process, and identify a vague image. On the other
hand, too fine a screen (such as 300, on some photographs)
may result in a dark, ink-clogged image in which the dots are
so run together that they cannot be resolved.
An experiment on lighted lines of dots (Indow and
Watanabe, 1984), recalling depth-perception and peripheral-
vision tests for a driver's license, provides insights for
preparing holograph illustrations as they become more
common. An editor can usefully study what movement and
three-dimensional information are perceived by the viewer
and how material should be arranged so that perception of
desired materials as well as relationships results. Using dots
as subject material for depth perception, as Indow and
Watanabe (1984) did, parallels the device of screening
information in complex and multilayered geotechnical
illustration. In most geotechnical map illustrations, many
layers of information must appear, yet the illustrations must
retain some transparency. The reader must be able to look
into an illustration, perceive its three dimensions (the
layering of information), perceive the screened information as
background (as behind the primary information both in depth
and in importance), distinguish the primary information

from all secondary information, and process the whole as an
integrated figure having depth and layers of emphasis. The
success of a complexly layered map illustration also depends
on the reader's ability to transfer her understanding of the
figure to an understanding of what occurs in the geographic
area that the map portrays.
Dots, especially those produced by screening, perform
an important and unavoidable role in producing three-
dimensional clarity in geotechnical illustration.
Psychophysical experiments indicate that dots are basic
building blocks of visual perception, and thereby suggest how
and why they can be so useful and successful in illustration.
How People React to and Perceive Lines in
Psychophysical Experiments
From the ways experimenters used lines in the
experiments in Perception & Psychophysics, people seem to
consider that lines express something a little closer to reality
than dots do. The researchers used lines to present more
definite, defined outlines and to test how quickly subjects
perceived, completed, and followed the direction of those
outlines. From the results, people construct more completed
and complex percepts from lines than from dots. Lines
attract people's organizing and completing capacity more

than dots do. Thus, a contour (a line) seemingly presents
more two-dimensional reality to the viewer than a dot does,
and its movement (real or illusory) seems easier to perceive
and interpreted as closer to reality than that of a simpler
element such as a dot.
Jiao and others (1984) studied lines to examine three
issues: how information is transferred from eye to eye, how
much information moves in that transfer, and how fast it
moves. They found that, "Stationary lines appear to move
upwards following exposure to downward moving lines when
the two displays are viewed either with a single eye or with
both eyes" (p. 105). Our left and right eyes do not see quite
the same things, so the perception system must transfer
information about relative motion of lines back and forth (Jiao
and others, 1984):
Monocular and binocular exposure to opposite
directions of motion also yields aftereffects that
differ in their directional properties in
accordance with whether one eye or both eyes are
used in viewing the stationary test display.
Neither effect can be explained by assuming that
binocular processing simply involves an additive
combination of inputs from the two monocular
channels (p. 108).
Bagnara, Simion, and Umilta (1984) tested subjects
ability to recognize variously oriented pairs of letters as "same
or different." To test specific reactions, they varied their
experiments in some cases by framing a test letter in a

triangle made of lines. The line frame seemed to accelerate
responses, when the response was that both letters in the
pair were "same." Lines or combinations of lines juxtaposed
to symbols or letters in graphic material seem to enable us to
perceive, process, and take the message more efficiently
from the graphic material; the lines point the way to the
symbols that we need to interpret (Bagnara, Simion, and
Umilta, 1984).
According to Zucker, Stevens, and Sander's (1983)
work with dots, the "visual system" does "tend to organize
into contours those dots that are adjacent, similar, and
(roughly) collinear" (p. 515). People also seem to readily
recognize, organize, and follow linessolid or screenedin
graphic materials. People are strongly influenced by contours
or lines that appear in fields that seem to contain
homogeneous stimulation (Petry and others, 1983): for
example, they can be distracted by an illusion of boundaries
or contours in an area of continuous pattern or clustered
symbol. Such subjective contours or illusory boundaries are
so compelling that material within them (such as symbols or
patterns) can appear brighter than that outside them (Petry
and others, 1983). Because we are predisposed to
completing contours (for example, we want to string adjacent
dots together [Zucker, Stevens, and Sander, 1983]), and
because we are drawn to lines as outliners and pointers, we

can be confused by them (Petry and others, 1983) if they are
Finally, people misperceive the orientation, the relative
orientation, and the direction of lines. For example, we do
not automatically judge correctly the center of concentric
arcs of a circle "drawn in different circular sectors" (Roncato,
1983, p. 43; example, Appendix A); we tend to "under-
estimate the curvature of short arcs and consequently to
misperceive their centers" (Roncato, 1983, p. 44). In our
attempts to gauge the relationships between curved lines, we
mentally construct end tangents coming out of arc segments.
(An end tangent is a straight line extending from the very end
of a curved line segment, its direction depending on the
direction of that very end of the curved line.) Then we
estimateusually unsuccessfullytheir parallel or nonparallel
nature in order to judge their concentricity. Or we
underestimate curvature of a short arc and therefore
overestimate the corresponding circle's radius (Roncato,
1983). To perceive concentricity with any exactness, we
must satisfy both "end-tangents parallelism and centers
coincidence" (Roncato, 1983, p. 52). It seems that when
observing lines we often look back and forth between
segments to judge their relative position, something like the
way we look at the line of near hills, that of distant hills, and
the far horizon and estimate their position relative to each

other and to us. But we cannot consistently gauge the
relation of different line segments to one another, even when
there is a precise relation there to be gained (such as
concentricity of arc segments); we are not good at
automatically seeing it.
Because research subjects used lines to complete
percepts of outlines and shapes, lines seem to present
something closer to reality than dots. Using lines,
researchers tested perception of linear movement, location,
relative position, quantity, and directionality. Lines provide
this information very commonly in geotechnical illustrations.
Lines as Important Elements of Geotechnical Illustration
Geotechnical illustration does not exist without lines.
Therefore, if lines in research experiments enhance subjects'
perception or in some cases distract it, line illustrations may
elicit analogous responses.
In geotechnical illustrations, lines (called "neatlinesj
box in the illustration material, and show the viewer the
visual limits of illustrations' information. On maps, lines
perform many important functions: they separate different
geologic units, locate structures such as faults, outline
mountain ranges, and show the culture items such as roads
and rivers. On graphs, lines form the axes, the curves within
the graphs that compare the axes' values, and the boxes that

present values on histograms. Lines covered by a screen (in
drafting or printing) and changed to gray lines composed of
dots present background information (often the culture
information in a map) upon which lie solid lines presenting
the primary information (the geology); using different
darkness, intensity, or weight of lines allows more kinds of
information to be retained and differentiated. Petry and
others' (1983) definition of a contour, "an abrupt change in
luminance, wavelength, or purity between adjacent regions in
a stimulus array" (p. 169), reinforces the importance of lines
in perception and subsequent comprehension. An abrupt
change marked by/represented by a continuous stimulus such
as a line definitely attracts attention. Lines draw the reader's
attention in geotechnical illustrations as they seem to do in
psychophysical testing; and this attention-getting, though
vital in illustration, can sometimes cause problems for
perception. For example, an illustration may employ several
patterns composed of distinct lines, in one pattern strongly
vertical, in another horizontal, in another at an angle. One's
vision is led in several directions at once, which causes
confusion in integrating the illustration. Each eye may be
trying to process one or more of these directions and transfer
information to the other during perception (as in Jiao and
others, 1984). What each eye sees is difficult to integrate
because of perceived (illusory) motion.

A very useful and standard device in geotechnical
illustration is using lines to "frame" type: draftors commonly
arrange the letters of a river name to curve along the line
representing the river location, an economical use of space.
According to Bagnara and others' experiments (1984), such
framing may also make the letters easier to read. We
apprehend three things at oncethe river's presence, the
river location relative to other features, and the river name
efficiently processing all these components simultaneously.
Solid-line geologic contacts, separating different geologic
units, probably act as frames: a unit delineated (framed) by a
line boundary seems easier to recognize and integrate with
other information than a unit identified with a pattern or
name but no line boundary (the pattern ends within a space
the rest of which is left blank). Using straight-line segments
to connect names with geologic features or to connect labels
with curves on graphs also should quicken perception and
comprehension, because the identifier label is framed within
its location space and against what it identifies.
The presence of subjective contours, or more
specifically illusory boundaries (Petry and others, 1983), in
the form of lines, may underlie the failure of some
geotechnical illustrations. (A contour marks a gradient within
visual material representing the same thing, such as
topography, and a subjective contour is one that is illusory. A

boundary separates different meaning fields, and an illusory
boundary is one that is perceived but is not meaningful.)
Similar line weights drafted for different kinds of information
(topographic contours, roads and rivers, geologic contacts
between units) distract the eye, introduce illusions, and block
comprehension, because the reader cannot quickly separate
what lines identify what features. Although I found no
research specifically testing perception of different line
weights, Petry and others' work on subjective contours
supports a necessary technique in drafting. That is, effective
line drafting varies line thicknesses, uses screening to vary
intensity (but maintain continuity of the contour), and
alternates continuous line and dashed line, all of which lessen
the reader's chances of getting lost among many kinds of
linear information. And one of the ways people get lost in
complex line information is to perceive illusory boundaries or
subjective contoursline information that does not really
Roncato's research (1983) on perceiving concentricity
of arc segments also suggests why readers struggle to follow
map contours (such as topographic contours or geophysical
contours) where every contour cannot be shown as
continuous because they are too close together for the map
scale in a particular area. Trying to follow and connect these
discontinuous curving contour segments may imply that the

reader must first judge concentricity of the contours and
then locate one or more centers of the arc segments.
Similarly, graphs using arc-shaped line segments that overlap
and cross each other may be difficult to sort and process
partly because of similar implicit problemsdetermining one
or more orienting centers. Both the research experiments
and practical aspects of drafting suggest that varying the line
weights, making some lines dashed or dotted, coloring some,
or screening some serves to sort the lines, makes them
easier to locate and more effective to process, and may defuse
underlying potential anxiety about centers; if the lines are
strongly different visually, processing may not have to resort
to differentiating centers to get started.
The usefulness of lines in geotechnical illustration for
separating features and for showing their location and
directionality, and the problems in sorting and judging the
relative depth position and importance of lines all have some
background in visual perception, as suggested by the
experiments in psychophysics.
How People React to Patterns in Psychophysical Experiments
Patterns, which are repeating graphic elements in a
specific arrangement in a two-dimensional field, provide
researchers with a potentially more complex and physically

larger visual area in which to test human perception.
Psychophysical researchers wanted to learn the following: (1)
how people perceive patterns, (2) how fast patterned areas
are integrated by the brain through the visual system, (3)
where in the brain this occurs, and (4) what pattern
combinations are quickly perceived and what ones are
distracting or illusory. They found that people readily
differentiate patterns from and against a background, but they
react differently to different intensities of pattern versus
background. They perceive some pattern combinations to
show up clearly and to be easy to perceive and comprehend,
whereas others blend together and are not clear. Still other
combinations present such stark contrasts as to be
perceptually distracting, and readers cannot integrate them.
Contrast and assimilation of black and white patterns
are of general interest in psychophysical research. This
research clarifies what constitutes contrast versus
assimilation; how they are perceived (how and why people
differentiate patterns or mix them together); and where
contrast or assimilation registers in the brain (for example,
Hamada, 1984). Perception of contrast also influences
perception of depth and of closeness of one graphic layer or
another to the viewer. Brown and Weisstein (1988) noted,

If regions filled with relatively higher spatial
frequency sinusoidal gratings are adjacent to
regions containing relatively lower spatial
frequency gratings, the regions with the higher
frequency will appear closer in depth than those
containing the lower frequency (p. 157).
Not only dots and line contours can cause people to quickly
complete a figure from incomplete stimuli (Uttal and others,
1988; Zucker, Stevens, and Sander, 1983). Pattern areas can
also stimulate perceptual completion of a figure, and the
completed figure may be a misleading, distracting illusion,
radiating stronger "apparent brightness" and "apparent
sharpness" than surrounding material (Petry and others,
1983, p. 169).
We exert our perception capacity to complete figures
and especially to "perceive a stable visual world despite the
fact that we make discrete eye movements every 250-300
msec, on the average" (Rayner and Pollatsek, 1983, p. 39).
We integrate these successive views, blocking out blurs
between focuses; we do not get confused by them, and we
maintain our constant perception stream. The brain may
integrate successive views and block out saccades (blurs
between focuses) using a memory buffer that aligns and
integrates information across saccades (Rayner and Pollatsek,
1983). In Rayner and Pollatsek's experiments, subjects
viewed a pattern of 12 dots located off their focus center, and
when a subject's vision jumped toward the pattern, 12 other
dots were substituted. The question was asked, "Which dot of

the two 12-dot patterns changed place?" Because the results
were far above chance occurrence, the conclusion was that
clearly "a short-term memory buffer [exists] that captures
much of the geometric reality of the visual scene that we see
by means of a series of fixations" (1983, p. 47).
Another kind of completion or integration of pattern
material can be repeated from the earlier discussion of dots.
People also readily recognize both the symmetry and the
textural equality that result when a dot-texture (a group of
dots whose image taken together also constitutes a pattern) is
repeated identically across a line axis (Jenkins, 1983)--
something like recognizing the symmetry and textural or
pattern equality of butterfly wings regarded across the axis of
the body.
People look for edges and centers (particularly centers)
of pattern fields. Proffitt, Thomas, and O'Brien (1983) tested
that drawn interest of people to edges and centers by having
subjects observe three lights on the edge of a rolling wheel
(in darkness). The subjects perceived a centroid for the
three lights and oriented their understanding of movement
around that centroid. They did not perceive the rolling
wheel motion; they were determined to perceive centers and
boundaries and to ignore the rolling motion. Proffitt,
Thomas, and O'Brien then stated, "The visual system does
not equally apprehend all points within bounded shapes, but,

rather, certain locations are of special importance for
perception" (p. 63). As people look for "centers first," so they
seem to first perceive patterns from a few composing
elements (such as dots) and then proceed outward to
integrate more elements (such as dots) into the percept
(Zucker, Stevens, and Sander, 1983). These two
experimental situations seem to comment on each other:
wherever those "first few dots" are may define the center that
we begin with in processing a dot shade which becomes or is
perceived as a pattern.
Both the emphasis on brightness, and the illusory
nature of the experimental material observed by Proffitt,
Thomas, and O'Brien (1983), recall the experiments of Petry
and others (1983), who found in testing illusory fields and
subjective contours that "apparent brightness is influenced
more by number of inducing elements, whereas apparent
sharpness increases more with inducing element width" (p.
169) and "perceived brightness and sharpness are two semi-
independent components of subjective contours" (p. 174).
The point is that brightness and sharpness and illusion
influenced subjects looking at the pattern made by lights on a
rolling wheel, as well as subjects looking at black and white
patterns as such. Brown and Weisstein (1988), presenting
subjects with higher and lower spatial frequency black and
white gratings, also observed such effects of brightness and

sharpness and found that varying these does make a
difference in perception, especially that of relative closeness
perceived by the viewer of one pattern versus another.
People have some difficulty judging relative sizes of area
and volume, a task which a reader may wish to do by
estimating areas of pattern material on a map. For
experimental subjects,
area and volume judgments are usually
overestimates of true size. Subjects estimate
linear distances reasonably well, but error
escalates as mental multiplication is done to find
area and volume (Butler and Overshiner, 1983, p.
Rayner and Pollatsek (1983) noted that we "perceive a stable
visual world" even though our vision naturally blurs
momentarily between focuses. However, Butler and
Overshiner (1983) suggest that we do not always perceive
that visual scene accurately. Inaccuracies in judging the size
of elements of the visual scene are "not simply the result of
rounding followed by accurate multiplication" (Butler and
Overshiner, 1983, p. 597).
Lines were used in perception experiments to study
perception of movement, directionality, outline, and contour.
Patterns were used more to study perception of areas, that is,
how we integrate and catch the idea of an area (that it exists
and that it is to be perceived together, that the two-
dimensional area rather than a dot or a line provides the

percept), after which we may proceed to process its sense.
The repetitive nature of the pattern across two-dimensional
space stimulates that ability to perceive an area as one
percept. And the necessity to perceive one area rapidly as a
clearly defined whole is extremely important in geotechnical
illustrations. Such an area also needs to be clearly compared
with or set off against other neighboring areas, and not
confused with other areas having different meaning.
Information that suggests how people perceive patterns and
what problems may result from their misuse can be useful for
preparing effective geotechnical illustrations.
Patterns as Important Elements of Geotechnical Illustration
Patterns are extremely common in geotechnical
illustration, especially in maps but also on graph forms. (In
histograms, for instance, where sizes [heights, widths] of
blocks define values and allow for their comparison, patterns
make the values easier to read.) Patterns define specific
areas that are important to perceive as two-dimensional
Patterns on a map allow the reader to quickly
distinguish among the geologic units (rocks or sediments)
represented. They often show the character of the unit as
well, a classic example being fine dot patterns traditionally
used to indicate the graininess and bedding forms of

sandstone (example, Appendix A). Patterns may also be used
to define the area of a structural or a physiographic basin or
uplift. In addition to illustrating geology as such, the use of
patterns may separate mountainous areas from valleys, or
point out swamps, lakes, and forests, thereby providing
background information about an area, to indicate
accessibility for exploration.
One principle of effective pattern use in geotechnical
illustration is that patterns should be quickly differentiable.
Jenkins (1983) found that people recognize symmetry and
equality quickly if a pattern presented to their visual field is
rotated across a line axis and repeated, and his findings
suggest the reasoning behind the use of noticeably different
patterns for geologic units. If the same pattern appears on
both sides of a line drafted to show units' separation in the
field, the reader can easily follow a misleading perception
process: she can perceive that symmetry across a boundary,
process the symmetry as equality, treat the two units as
lithologically equal, and consequently misunderstand the
geology of the mapped area.
Another drafting principle for applying patterns is to
find, represent, simplify, and consistently use a visual relation
between pattern and geologic unit characteristics. Subjects
in Zucker, Stevens, and Sander's work (1983) studied an
evenly-gridded dot-rectangle pattern, a "linear hash," which

3 8
is identical to the pattern for "massive sandstone" used on
geologic maps. That they considered such a pattern as basic
to their research suggests the usefulness of this pattern for
representing geology: both experimental subjects and
illustration readers perceive such a repeated evenly-gridded
pattern smoothly, beginning with perhaps two dots and
extending perception outward. With the added context, that
of sandstone, readers add a memory connection, dots
standing for grains; readers thereby distinguish massive
sandstones (evenly gridded dot patterns) from crossbedded
sandstones (bands of dot pattern within each of which
parallel lines of dots run at an angle). They also learn to
connect other pattern combinations specifically with other
kinds of rocks. Our perception, as well as what we have
stored in memory and are primed to recall, is obviously
important for recognition of dot patterns; conversely, that we
recognize dot patterns as being meaningful is important to
our perception.
Draftors preparing illustrations with patterns as a rule
avoid using patterns that draw the eye strongly in different
directions--patterns with heavy-contrast line or symbol
directions that distract. Where such patterns are used, they
lead the reader's vision in several, distracting directions, and
in Jiao and others terms (1984), each eye may be trying to
process one or more of these directions and transfer

information to the other during perception. Perceiving these
linear motions must block or slow integration of the percept
of one patterned area. The distraction in such linear-
directional patterns probably thus makes it more difficult to
apprehend and accurately estimate total area or volume of an
area, which Butler and Overshiner (1983) found to be difficult
for subjects. These patterns thus confuse or negate one
important function of patterns on geotechnical illustrations
such as maps, which is to help the reader (1) apprehend the
entire area, (2) pull it together in her mind, (3) make a
usefully accurate estimate of its size, and (4) transmute that
size to a concept of the geographic area.
When draftors choose patterns to convey geologic
information or when they choose screens to produce various
intensity levels in layered geologic information, they address
issues of pattern contrast and assimilation, which are the
subject of general research by Hamada (1984). Geotechnical
illustrations communicate poorly if they contain too glaring
contrast or too little contrast (1) between pattern and
adjoining pattern, (2) between emphasized pattern in one
area of an illustration and a background shade in another, (3)
between pattern and background shade in ONE area, or (4)
between pattern and selected screen of underlying
topographic base. (A book full of such illustrations is tiring
and frustrating to process, and in the final analysis it will not

4 0
be read.) Further, discrimination among gray shades A, B,
and C, when all are on one illustration or on several figures
arranged together, can be difficult if the shades are too
similar. A common reason for redrafting is that originally-
selected gray shade patterns assimilate, because they are too
similar to differentiate.
Patterns that are narrow "black and white gratings"
(Brown and Weisstein, 1988) of considerable sharpness
(examples, Appendix A), used on a map illustration or a
histogram, lift off the plane of the figure. They cause a
particular map unit or histogram bar to come closer to the
viewer, and to become difficult to comprehend, as Brown and
Weisstein's research with higher-frequency versus lower-
frequency grating patterns would predict. This perceived
closeness creates an illusion, which confuses the layering of
the illustration because each layer should be on one and the
same plane for clear processing. A histogram, consisting of
only one layer of information, might successfully contain
several gratings, with horizontal, vertical, and angled black
and white bars. Viewers might avoid the depth perception
problem in this situation because all sharp-contrast patterns
would operate on the same close-to-viewer level. On a map,
however, where all such patterns adjoin, where boundaries
between patterns are not parallel, and where other layers of
information must be read and integrated, using all bright

grating patterns is not effective. The patterns rise toward the
reader and compete with each other for importance. The
screened base information recedes in comparison, and the
middle-emphasis information, such as lines of structure
(faults, anticlines, synclines, in solid black lines) becomes
ambiguous as to when and how the reader is to perceive it.
The combination can produce such disparity in the reader's
perceived distance to the patterns and distance to the base
information that the illustration becomes disjointed,
discontinuously layered, and hard to integrate and process.
In some map illustrations, the patterned geologic units,
in an unusual role, provide the base. They are screened into
the background, and the important information (such as
complex structures or sample localities) is layered on top of
them. But even when they are screened, sharp-grating black
and white patterns used on the geologic units can still
dominate the illustration at the expense of the information
designated as important: Brown and Weisstein (1988) found
in many experiments that a higher frequency area was seen as
"figure" (the important information) and lower frequency area
was seen as "ground" (less important). That Brown and
Weisstein's bright gratings (patterns) lifted off the normal
viewing plane of a visual surface constitutes an illusion, which
creates and defines an illusory boundaiy. Petry and others
(1983) discussed illusory boundaries that appear in visual

element GROUPS (more like symbols, to be discussed in the
next section). The brightness and sometimes undesirable
attractiveness of these gratings, which approach the viewer
without that intent being built into their use, seem analogous
to the brightness and sharpness of Petry and others' illusory
boundaries around symbol groups. These gratings present an
unwanted third dimension, analogous to the third dimension
intentionally built into gridded computer-produced diagrams
that show peaks and valleys in data by expanding and
contracting a network of lines (example. Appendix A).
Readers of illustrations sometimes become confused in
looking from one pattern to an adjoining one because the
patterns seem too much alike to differentiate quickly, or
because something about one or more of the patterns is
distracting, such as a depth inconsistency or illusion. Editors
are aware of this potential confusion: they intuitively
understand and work from a determinable level of pattern
discrimination and differentiation. Such care in pattern
selection minimizes confusion in the memory buffer that
Rayner and Pollatsek (1983) believe people use to retain
material in memory during vision jumps, and which readers
may utilize while reading within illustrations.
Research in perception, as well as practical experience,
shows that pattern use in illustrations is seemingly more
complex than that of lines, so that it necessitates more

complex perceptual processing. Readers may follow lines
successfully or confuse which line is the river and which the
geologic boundary, and they may either find or miss the
location of a dot or of a dotted contour; but perceiving
patterns involves integrating more area, both while
perceiving all pieces of one pattern and while integrating the
extent and function of pattern A, with pattern B, and with
pattern ...n. Therefore there is an even greater potential for
confusion than occurs with dots or lines. The effective use of
patterns for identification and integration of two-dimensional
areas seems illuminated by the way people respond to
patterns in experiments.
Symbols and Tvpe--Recognition and Positioning
At first glance, symbols seem "simpler" than patterns
because they are individual characters; why treat them after
patterns and with letters, and not -with dots, if the
progression is from "simpler" to more "complex" visual
materials? The answer is a compound one. Dots can function
as symbols, because dots often are used in illustration to
locate towns, for example, and dots and other symbols (such
as plus signs, X's, solid and open triangles and squares) do
sometimes overlap in their uses and amount of meaning
transmitted. But that non-dot symbols are a more complex
matter of perception than dots is best shown by the use of

dots in screened material. Material that is screened to allow
the backgrounding of secondary but useful information shows
up as non-meaningful dots, not as other-shaped symbols,
which would seem to promise other meaning even when
grayed. In comparison to patterns, individual symbols seem
to convey more meaning than patterns as well, certainly more
than the regularly-arranged dots or crosses or little v's that
make up a pattern. A pattern is a visual guide to integrating
an entire area and separating it from other areas patterned
differently; a symbol directs the reader to a location from
which she can, for instance, interpret a value. From readings
in perception, non-dot symbols also present some of the
same recognition problems that type letters do, and their use
can allow visual illusions that are more difficult to eliminate
than those that accompany dots.
How People React to Symbols and Type Letters Used in
Psychophysical Experiments
Literate people look for letters, and they look for
symbols that are intended to convey meaning or provide
information. Since people readily give their attention and
interest to letters and symbols, experimenters have exploited
that interest in both recognizing and capturing the meaning
of letters and symbols to study such things as reaction time to
letter identification.

Psychophysical researchers wanted to learn whether
subjects' reaction time to locate and recognize letters and
symbols was affected by such things as: (1) placement and
angle of a test element, (2) size, and (or) heaviness,
thickness, and intensity of the element, and (3) spacing of
elements in the visual field. They also examined the process
by which people perceive these elements, whether people
fragment the elements and recompose them during
perception or perceive them by some whole. Various
separate experiments showed that all these factors affected
reaction time, though no experiment specified which acts
first or most importantly to affect the perception process.
Spacing may be especially important, however: occluded,
crowded visual elements were hard for subjects to begin to
process, and had the potential to present illusions.
Perceiving complex elements seems to involve both whole-to-
detail and detail-to-whole processing, operating in parallel.
Subjects tested for their ability to recognize differently
oriented pairs of letters as "same or different" identified the
sameness more quickly if two test letters appeared at the
same angle on the viewing area, though separated by space,
than if the presented letters lay at different angles (Bagnara,
Simion, and Umilta, 1984). In a variation on this
experiment, the upper letter in the view area was unframed
and upright, while the lower had a black box frame, a triangle

that could take one of six different angles but in which in any
case the top of the letter was in line with the top of the
triangle; in another experiment, both letters were framed,
and orientation of both letters and boxes varied. (See
Appendix A.) In all these framing experiments, the frame
seemed to accelerate response, when the response was "same
letter." Bagnara, Simion, and Umilta believed that
recognizing a pair of letters as same or different requires two
brain mechanisms or functions that seem to accomplish the
character transformation that we perform to check
"sameness": one is an actual mental transform, the other is
more automatic, without awareness, using what the
experimenters called a "visual code."
Some experimental evidence suggests that people
experience difficulty perceiving one letter within an array of
letters. Johnson and Blum (1988) noted some evidence that
it "takes longer to detect the initial letter of a redundant
array such as BBBBB than it does to detect a single letter
presented in isolation" (p. 147). Johnson (1986) argued that
subjects detect "presence or absence of a target letter" faster
if the next display is one letter than if the new target is part
of a "redundant display" such as BBBBB (p. 93), because in a
redundant display the subject tries to "encode the pattern as
a unit" (p. 94) before processing it. People seem to need a
certain amount of space between type elements to decide

that they are to be comprehended one-by-one rather than as a
People also differentiate at an early processing level
different categories of symbols. Whereas Johnson (1986) and
Johnson and Blum (1988) investigated reaction time to
detect letters within groups as opposed to a letter by itself,
Cardosi (1986) experimented with finding a target among
distractors of another category (letters versus numbers).
Cardosi's "alphanumeric category effect refers to the
finding that letters and digits are identified more efficiently
when presented among items from the opposite category
[letter versus digit] than among items from the same
category ." (p. 317). When the physical similarity of targets
and distractors was increased, subjects found locating a target
more difficult, even if the surrounding elements were
technically of the opposite category. It was difficult to locate
a C among a series of 8's, 6's, and 9's.
The amount of space between a target and a mask (an
element that hides the target or distracts the viewer from the
target), between the letter or digit to be identified and those
among which it lies, is important to perception (Wolford and
Chambers, 1983). When the target item is closely
"surrounded by other items" (p. 129), the "probability of
correctly identifying a target (e.g., a letter) is generally
reduced" (p. 129), but when the space is increased, confusion

is reduced. Wolford and Chambers noted that two nearby
items "are presumed to compete for a limited set of feature
detectors" on the retina (p. 129). At a close spacing of target
and mask, this "feature interaction" may dominate our ability
to group data. Wolford and Chambers concluded that more
work needed to be done on how the parts of the retina react
to interacting stimuli, because, for example, lateral areas of
the retina react differently than central parts to masking of
target stimuli by nearby masking elements.
There are optimum viewing distances for optimum
perception, which have been tested using letters (Raymond,
1986). Raymond's experiment tested subjects reaction time
for recognition of a small letters being replaced by another
during the viewing interval: problems with visual
accommodation, even if small, slowed reaction time.
Raymond noted: "Johnson (1976) found that accommodation
was most accurate when stimuli were viewed at a distance
that corresponded to the individual's resting position" (p.
281). Actual viewing distance to a target element is a
physical variable that influences the eye's efficiency in
reacting to that element.
Letters and symbols are the subject material for
experiments in how people process imagesexperiments
that test whether we perceive whole images followed by their
details, or instead perceive part by part to the whole (Lamb

and Robertson, 1988). In studying large letters made up of
small letter symbols, Lamb and Robertson (1988) noted that
some researchers "have agreed that global and local
processing can occur in parallel or at least proceed with a
similar time course" (p. 173). In their experiments,
Central presentation [to the fovea, the retinal area
of greatest acuity] decreased reaction times for
identifying small letters presented within a
hierarchical stimulus pattern (i.e., local letters)
but not for a single small letter presented alone
(p. 172).
Earlier research, said Lamb and Robertson, has shown that
subjects process a large letter "S" made up of small letter "S"
symbols, or a large letter "H" made up of small letter "H's"
much faster than an "H" made up of little "S's" or a great "S"
made up of little "H's" (1988, p. 172). The problem was one
of "Stroop interference," of which a classic example is to ask
a subject to call out the color of the word "green" when it is
printed in red. Stroop interference in this color example
comes about because we are primed by our culturally-derived
emphasis on reading to read a WORD, and must force
ourselves to perform another task with that word. If the task
contradicts the information of the word, Stroop interference
results. From Lamb and Robertson's testing with large and
small letters, there seem to be certain subtle physical and
psychological relationshipsperhaps many of themthat must
be balanced between parts of a visual element to be perceived

and the whole element, for perception of both parts and
whole to take place efficiently.
That we do not automatically judge correctly the center
of concentric arcs of a circle "drawn in different circular
sectors" (Roncato, 1983, p. 43) is a symbol issue as well as a
line issue. We may perceive potential end tangents projecting
from arc segments (designed to be seen as overlapping
symbols) to be either parallel or nonparallel as we decide on
the arcs' concentricity; or we may underestimate curvature of
a short arc and therefore overestimate the circle's radius.
And according to Roncato, we must complete both tasks
satisfactorily to judge concentricity of circles, that is, we
must satisfy both "end-tangents parallelism and centers
coincidence" (p. 52). This task seems to imply the need for
considerable processing capacity in studying groups of
incomplete-circle symbols. Our internal vision of the entirety
of any incomplete circle symbol is uncertain and faulty, which
can cause us to misjudge sizes within a graphic display. This
confusion can slow processing as well as cause us to
incorrectly interpret the display.
We are also misled about sizes of symbols, such as
circles, because we judge size of circle by the size of nearby
circles, and thus sometimes misjudge them all (Weintraub
and Schneck, 1986). Weintraub and Schneck's testing
combined a search for a target element among masks or

distractors with the observation that subjects misjudge sizes
in concentric and nonconcentric circles because of the
"context supplied" by distractor circles. They used two
famous circle illusions as test materials. In their results, they
noted that contours close to a test circle attract different,
farther-away contours, and that "large nonconcentric
inducing circles" make a test circle seem smaller (p. 155).
As Weintraub and Schneck summarized, "The contour/circle-
context model asserts that contours attract and that context
supplied by the larger or smaller size of nearby nonconcentric
circles leads to size contrast" (p. 147). Again, the fact that
larger and smaller circles, complete or incomplete, are
selected to appear in a graphic display does not guarantee
that all sizes will be clearly processed.
People's interest in perceiving centers of figures
(Proffitt, Thomas, and O'Brien, 1983) also may contribute to
success or failure of estimating symbol sizes, especially when,
as in the illusions tested by Weintraub and Schneck (1983),
symbols are closely juxtaposed: "Certain locations [in
bounded shapes] are of special importance for perception" (p.
Lastly, perception of symbols in groups is influenced by
the presence of subjective contours, illusions that tell us that
a shape is to be perceived as one image or percept when it
actually should be separated into separate meaningful

elements. Apparent brightness and sharpness of an image
field are variables that attract attention and influence
perception of a subjective contour. Petry and others (1983)
found that "apparent brightness is influenced more by
number of inducing elements, whereas apparent sharpness
increases more with inducing element width" (p. 169).
How subjects' perception mechanisms checked out
what letters and symbols were doing in visual displays was
what interested researchers in letter and symbol
experiments, and what they wanted to know seems to differ
from what other researchers wanted to learn from
experiments with dots. Experimenters with letters and
symbols wanted subjects to "identify targets," rather than
"watch dots," "watch lines," or "integrate the spread of
patterns across an area." They seemed interested in how and
at what speed subjects recognize symbols and letters,
integrate their positions, group ones that go together and
separate ones that do not, and understand them. Recognizing
a letter change and calling "same or different letter" are tasks
that seem to represent more complex perception, and seem
to require more judgment, than does recognizing lines of
dots or even recognizing the symmetry of a dot pattern when
it is repeated across a line contour.
Processing symbols and type in graphic displays
obviously is more than a matter of merely perceiving them,

and then deciding what they meanwhether from memory,
from ability to recognize and read letter groups, or from
explanations given with the display.
How Geotechnical Illustrations Use Symbols and Type
Illustrations in geology use a multitude of type letters to
identify axes of graphs, to identify particular symbols on
graphs or the columns of histograms, and particularly to label
geologic units, structures, and "culture" (roads, rivers, towns,
parks, forests, mountain ranges, and so on) on maps. They
use symbols (non-alphabetic-letter visual elements such as
X's, squares, circles, triangles, plus signs) to locate points of
importance, such as a locality where a stream-sediment
sample tested out at 0.05 ppm gold. One map or a sequence
of maps of an area may use many such symbols to pinpoint
sample localities tested for many elements or minerals.
Placing type letters and symbols so that readers can
apprehend them and not get lost or confused is extremely
important to the usability of one illustration or a sequence of
related illustrations. Because type and symbols provide
higher level meaning to the illustration, it is both important
and easy for draftors to concentrate on clear placement.
Further, because drafting standards and practices do
emphasize selecting and placing these elements, many

perception issues suggested by the experimental readings are
simplified or resolved.
Type and symbols on maps, for example, name and help
to locate the geographic features; they pinpoint important
locations, such as of ">0.07 ppm gold in sample"; and they
tell us how to interpret what patterns represent on a map.
Because type and symbols convey so much of the higher-level
meaning of illustrations, draftors are extremely aware of them
and lay out their bits of material carefully. Draftors devote a
great deal of attention to spacing type and symbols so that
they do not overlap, either when placed on one layer of
information, or when one or more geologic information layers
are composited to a base information layer. When a draftor or
editor checks a drafted composite proof, "overprinting" of
one element on another is a problem that is immediately
looked for, but not often found because of this care in
drafting. Draftors place a corresponding amount of care in
placing type and symbols so that they clearly refer to the
intended area.
Thus, a standard rule in drafting supports what Bagnara,
Simion, and Umilta's testing (1984) suggested; that subjects
are aided in recognizing type if target letters appear at the
same angle, and also if guiding lines are nearby. Draftors
curve type for stream names along the stream lines, angle
type names along roads, and spread them out through the

5 5
area of a mountain rangewith attention to spacing names
evenly, and at the same or only gently deviating curving
angles. When an area on a map already contains considerable
visual material (lines, patterns, type), draftors add
unobtrusive but clear "leader lines" to clearly point type or
symbols to a location.
Geotechnical drafting also automatically addresses the
problem of detecting "the initial letter of a redundant array
[of letters]" (Johnson, 1986; Johnson and Blum, 1988).
Redundant arrays in graphs and maps are almost always
represented by non-alphabetic-letter symbols. The use of
such symbols seems to simplify sorting and processing. The
viewer perceives that the symbols represent some meaning to
be interpreted, but she does not specifically try to complete a
unit (compose a word) out of the symbols, as she would
attempt to do if letters were used.
Drafting standards also correspond to the conclusions
of Raymond (1986), that optimum viewing distances exist
that bring about optimum perception. Particular sizes of type
and symbols (as well as particular widths of lines and weights
of patterns) are standard, and deviations from them (sizes
and weights that overpower the illustration or those that the
reader must squint to see) seem immediately noticeable and
are immediately marked for change.

Drafting standards also seem to address the processing
problem observed by Cardosi (1986), that is, locating letters
and symbols "among items from the same category" (p. 317).
Draftors select from among serif roman and italic type, sans
serif roman and italic type, and bold and light type, to
minimize such problems and make differentiation clear.
They also use these selections consistently: for example,
stream names are always serif italic nonbold type; rock unit
letter symbols (Qa for Quaternary alluvium, for example) are
always sans serif roman; town names are always serif roman
nonbold. As noted before, drafting standards also require a
careful use of spaceleaving as much space between items as
possible for clarity, using leader lines in busy areas, and
avoiding overprinting.
Problems for drafting and reading type and symbols in
geotechnical illustrations do occur, however. Type and
symbols CAN be placed poorly, sized incorrectly, and jumbled
together in the illustration space. But one complex matter of
perception seems to revolve around several experimental
issues: (1) whether we perceive visual material as a whole,
then perceive details, rather than perceiving part-by-part to
the whole (Lamb and Robertson, 1988); (2) why we misjudge
symbol sizes and cannot always determine centers (Proffitt,
Thomas, and O'Brien, 1983; Roncato, 1983; Weintraub and
Schneck, 1986); and (3) what particular illusory problems

can accompany BOTH of these tasks (Petry and others, 1983).
Petry and others (1983) concept of illusory boundaries and
subjective contours seems the overriding issue. Symbols (in
particular) often appear on geotechnical maps in groups,
usually locating areas of concentration of particular
commodity elements (gold, lead, copper, manganese)
searched for. These symbol groups, usually the brightest,
most attention-attractive bits of an illustration, can take on an
outline of their own, which is confusing, and distracts the
viewer from apprehending the values of each symbol. The
problem worsens when several elements (represented by
different symbols, for example, solid triangles, solid squares)
are represented clustering around one location. The
resulting combined illusory outline can mix several values of
several elements, necessitating even more attention to escape
the illusion, and resulting in even slower processing of the
illustration's meaning. This basic illusory problem gives rise
to many specific complications: in a cluster of symbols, such
as circles, a viewer may misjudge sizes (and thereby values--
the meaning of the individual symbols) because nearby
complete symbols assimilate in size to the current target (a
result seen by Weintraub and Schneck, 1986) or because
nearby symbols are incomplete, or occluded by others. Thus
their curvature, radius, center, and size cannot be clearly
judged (an example of the problem of perceiving and

5 8
organizing incomplete arc segments, tested by Roncato,
1983). These perception problems may make it difficult for
the viewer to determine a center of a symbol group, in
particular, a spot to begin processing (a human concern that
Proffitt, Thomas, and O'Brien [1983] observed). In addition,
that the viewer cannot find a beginning center may make the
misjudgments of size more compelling and more difficult to
escape. Finally, if a viewer can neither sort individual
symbols' shapes and sizes quickly nor easily determine a
center from which to begin processing, she cannot easily
initiate either (1) the process of perceiving the whole, then
the details of the cluster, or (2) the process of perceiving the
details, then the wholethe two processes which Lamb and
Robertson (1988) believed probably occur in parallel.
Thus, that useful standards exist for drafting symbols
and type individually and in strings on geotechnical
illustrations does not eliminate perception problems or
negate psychophysical issues. Illustration problems and
successes parallel results of the use of the same devices in
perceptual experiments.
How People React to Color, from Psychophysical Experiments
Researchers were interested in how people see color,
react to color, change their idea of what color is, remember

color, and adjust their perception to locate color. They
regarded color perception as an integral part of how people
relate to their environment, and color as an extremely
informative part of the environment. In particular
researchers wanted to examine how discriminating people's
color vision ishow precisely people differentiate colors and
how important color qualities and differences are to them.
They wanted to learn whether these concerns or
expectations influence subjects' reaction time to search out
color targets in visual displays. Results showed (1) that
adjacent colors in a display can assimilate, that is, be seen as
looking like each other, (2) that spacing between color items
in a display influences reaction time, (3) that certain colors
are preferred from people's memories, and (4) that
perception of color changes with changes in people's physical
People seem to develop, possess, and be able to pick
both a remembered color and a color that they prefer for
common items such as bananas, for example (Siple and
Springer, 1983). Siple and Springer controlled their
experiment carefully. Experimenting with what they called
the recognized attribute of color memory and preference,
they "calibrated" subjects' perception to a white card until
their perception of "white" was measured; then the subjects
viewed silhouettes as well as gray-toned screened pictures of

fruits and vegetables and were asked to pick both their
remembered color and their preferred color for the items.
(Experimenters used a Munsell color chart, a chart for color
identification in which color chips present a numerically
defined range of hue, value, and chroma, to make the choices
measurable across subjects.) The subjects remembered and
preferred more saturation (more pigmentation, "richer"
coloration) in colors than was "really" in the actual items,
whereas hue and brightness they chose accurately.
Tsai and Lavie (1988) studied the processes by which
subjects could report additional targets of the same color as
one that they were first directed toward in a graphic display,
and where in the display those additional reported targets
were. They found that "additional letters reported tended to
be adjacent to the first reported target(s)" (p. 15). They
suggested that "the selective processing of targets specified
by color or by shape is accomplished by attending to the
targets' locations" (p. 15). Color can be a help to easier and
faster processing, it seems, and looking for a specific color is
important in focusing attention.
Subjects in research by Fach and Sharpe (1986)
demonstrated the principle of the "influence of two or more
adjoining chromatic fields on each others perceived color
and lightness" (p. 412). The subjects were to match
perceived hues of grating bars (alternating color stripe

combinations) with individual swatches of a 20-swatch
Munsell color set. Six grating combinations were used,
among the colors red, yellow, green, and blue. Red-blue pairs
assimilated: the subjects reported the colors more alike than
they actually were. Blue shifted toward yellow in the blue-
yellow combination: it was reported as more yellow than it
actually was; green shifted toward blue in that pair; yellow
always shifted the least. In considering perception of color
contrast versus assimilation, Fach and Sharpe built on some
historic research: they cited Chevreul, in the 1800's, as the
"first to quantify simultaneous color contrast (e.g., red placed
next to blue appears yellow)" (p. 412). Fach and Sharpe not
only used side-by-side combinations of colors but also used
color-bar widths that "varied from 2' to 20' [minutes] of visual
angle Hue shifts were largest for bar widths of 2';
however, they depended on the color combination used" (p.
412). Color next to color influences assimilation or shift, but
so does stripe or grating width. People are affected, in
perceiving a particular color, by the surrounding colors, and
thus, merely because adjoining colors are "different" in a
graphic display does not guarantee clear differentiation in
Human perception of color is not constant. Visual
function is modified by differences or changes in "visual
acuity, color vision, threshold sensitivity" and aging (Coren

and Hakstian, 1988). And specifically, color vision "may be
modified by environmental factors, exposure to
pharmacological agents, hormonal factors, or specific disease
states" (p. 115).
No research that I encountered compared reaction
time for subjects' perception of color targets as opposed to
line, pattern, type, or symbol targets. Given our culture's
insistence on color, however, and given that we stimulate our
children with bright color from infancy on, I would expect
color to be a primary, perhaps the primary, first-observed,
and first-looked-for aspect of a visual display.
Color as an Important Element of Geotechnical Illustration
Authors of geotechnical reports believe that color
makes their illustrations look more professional, more
complete, and more important. They consider color a vital
part of illustrating. Given any choice, authors prefer the
addition of at least one color and preferably a whole four-color
range to almost any illustration, although in many cases they
will consent to black and white for reasons of cost and
preparation time. Psychophysical experiments with color
suggested why color is so important and how draftors should
be aware of problems in illustration that may arise when color
is used.

The articles by Coren and Hakstian (1988) and Siple
and Springer (1983) suggest that color perception may be
modified by cultural change. What colors people look for and
prefer in art, house decoration, clothing, and technical
illustration vary with time, and doubtless vary geographically
and demographically across a society. That color preference
as well as perception varies with age is illustrated by the fact
that senior geologist authors often prefer bright, highly
saturated colors on their oversized geologic map plates,
whereas geologic drafting groups' taste in color over the last
15-20 years has evolved toward using mostly pastel, grayed
colors that simply seem less overpowering and that are
considered to be closer to "earth colors." In part, the
difference in preference by the older author is for "what used
to be done and what he/she is used to." Part of the disparity
may be color perception by older eyes: sensitivity to the
different wavelengths that produce the perception of color
changes with age. Older people's vision does not perceive
bright colors as being as bright as younger people's vision
does. Older authors may actually see pastels as more washed
outand thus less successful and worthwhile on their
important mapsthan the younger draftors do. Which color
scheme is better for the audience is a question: bright colors
attract, but a less-saturated pastel scheme is less tiring and
allows easier reading both of overlaid type and symbols and of

underlying base information. Often a drafting compromise is
reached by using bright colors on important units, especially
if they are small, and using softer colors throughout the
remainder of the map.
Color memory and preferencelearned preferences
that we recall (as tested by Siple and Springer, 1983)
influence color selection in geotechnical illustrations.
Authors, editors, and informed readers are offended by color
misregistration or color mistakes in multicolor materials.
Editors and authors comment passionately on proofs: "Color
washed out," "Too gray!" and "Check registration!" (when a
sliver of white occurs between composited colors), "Hope sky
color is more real in final printing," "Adjust color of red
rock," or "Is this really solid red?" All these readers have
distinct, strong expectations for color. Further, readers of
reports about geology learn to identify colors with geologic
ages: expecting yellow on Quaternary materials (from most
recent to 1 million years old), reds, roses, and oranges on
Tertiary materials (>1-66 million years old), greens and
khakis and browns on Cretaceous (>66-138 million years
old), blues to purples on Paleozoic rocks (240-570 million
years old), and so on. Because readers identify the rocks with
the color, the illustrations are not as coherent when colors
other than these learned ones are used. That readers take
such offense at color "errors" and color misregistration in

drafting represents partly a "perceived and preferred" color
situation and partly a preferred information separation or
boundaiy situation. That is, in some sense we take offense
because in both our memory and in our constant processing
of "reality," the same colors appear consistently where we
remember them"in the mountains all pine trees are green;
in southern Utah all Navajo Sandstone is golden reddish tan"-
-and no sliver of distracting white occurs between color areas.
The colors on rock units of maps also serve in part to clarify
the character and stratigraphic relations of rock units, and
thus appeal to another kind of memory, that of how the rocks
overlie each other in the field and what their sources are. For
example, the yellow applied to units of Quaternary materials
lifts off the page, which reminds us that these young surficial
materials lie on top of the bedrock. Red or orange drafted for
Tertiary materials recalls that these are very often volcanic
materials and were once burning hot, ejected or flowing from
volcanic vents or cones.
Similar concerns about exactness of color surround
printing of color photographs. Authors take great pain to
adjust four-color separations of photographs, so that just the
right balance results in colors that accurately reflect the
rocks depicted. They will even comment on the blue of the
sky, and ask for an adjustment in the cyan film separate, if
they believe that it does not reflect the right color and thus

6 6
incorrectly influences (assimilates with) the colors of the rest
of the photograph.
Considerable contrast is necessary to differentiate
adjoining color blocks on geologic map illustrations. The
principle of color correlated to rock-unit age is invoked for
any particular map or map series with the goal of (1)
maintaining noticeable differences both between rock units
and between large geologic time units, and (2) minimizing
confusing color assimilation, so that the geologic
relationships remain clear. When named units within age
categories proliferate, shades of color are selected to
differentiate them, and patterns in black and other colors can
be overlaid to sort the units further. For example, authors
differentiate Quaternary units such as landslide deposits,
alluvium, pediment deposits, terrace deposits, and colluvium;
and these terms and their deposits all need to be
distinguished by the reader, perhaps for practical purposes of
land-use planning in addition to understanding the research
for its own sake. If such varied deposits cannot be drafted
and printed in distinguishable color shades, they will not be
discriminated, whether they are in adjoining or separated
spots. The influence of adjoining colors on each other (Fach
and Sharpe, 1986), thus, is an important point in rendering
complex map illustrations understandable.

Tsai and Lavie (1988) suggested that looking for a
specific color is important in focusing attention; an adjunct to
this is that a specific targeted color attracts attention to a
particular area. Geologist authors know this color principle
well; for their color geologic maps they will persuade the
draftor to apply the brightest color currently in use to the
geologic units that they especially want to emphasize. On a
large fold-out map, for instance, the author may demand the
brightest color on a unit that is only an inch across on a map
area 2x3 feet, just so that one spot will catch the reader's eye
and signal her about what the author wants to tell her first
and foremost about geologic relationships.
Color on geotechnical illustrations is a desirable,
attractive, and useful addition. But both from practical
experience and from research findings, it must be applied
with discretion, so that different colors (and the information
they represent) can be discriminated clearly and so that
combinations are as pleasant to perceive and process as we
require them to be in our homes, art, or clothing.
How Psychophysical Test Subjects React to and Perceive
Deliberate or Accidental Illusions in Visual Material
Investigators of illusion and visual-processing problems
study not a positive reaction to or a direct response to

stimuli, but rather study what does not happen. That is, they
examine what causes reaction time to slow measurably, and
what in an experiment causes the subject to fail in
accomplishing the task. The specific concerns of the
research reported here are: the extent to which and
situations in which people (1) inappropriately complete
fragmented stimuli (and get the wrong percept as a result),
(2) are confused by illusory movement in stimuli, (3) are
susceptible to perceiving and being confused by subjective
contours or illusory boundaries, and (4) have problems
locating a target element among distractors. Researchers
were also interested in (5) how subtle or minor an illusory
stimulus can be and still mislead test subjects.
The most obvious and straightforward/simplest illusion
encountered in the psychophysical readings involves a
perception task which people succeed at every day, with very
useful results for their everyday living, and without realizing
that they have dealt with an illusion. People's vision readily
reconstructs percepts from incomplete stimuli (Uttal and
others, 1988). Uttal and others quoted research showing
observers viewing stimuli that were close
approximations of alphabetic characters
perceptually created stimulus components that
were not physically present, but were only
suggested by the arrangement of the other parts
of the stimulus (p. 223),

and they also stressed that, "The ability to perceive wholeness
in a world of naturally masked stimuli and noisy or
incomplete views of objects is obviously a very valuable skill"
(p. 223), In graphic materials, however, incomplete stimuli
can present a problem for processing information correctly.
Illusory movement in visual materials (Jiao and others,
1984) is also something that people are prone to pick up:
"Stationary lines appear to move upwards following exposure
to downward moving lines when the two displays are viewed
either with a single eye or with both eyes" (p. 105). Such
illusory movement in visual materials may lead the eyes in
confusing directions and slow processing, because the
directions do not have meaning.
The remaining research articles concerned with visual
illusion describe people's problems with subjective contours
or illusory boundaries and with locating a target element
among distractors.
A general definition of a contour is "an abrupt change in
luminance, wavelength, or purity between adjacent regions in
a stimulus array" (Petry and others, 1983, p. 169). In graphic
materials, contours are usually marked by lines, but can also
be marked by space between visual element groups or by a
change of abutting pattern areas. Petry and others' research,
however, noted that illusory boundaries (demarcations,
changes from one kind of information to another that the eye

perceives but that are deliberately not built into the material)
and subjective contours (gradations in same-kind information
that the eye perceives but that are not deliberate) can appear
in fields that seem to contain homogeneous visual stimulation
(Petiy and others, 1983). Material within these illusory
contoured areas can appear brighter (more important, closer
to the viewer) than that without. That illusory brightness
coupled with the illusion of the contour itself can attract the
viewer's attention, greatly mislead her by giving her pseudo-
information, and keep her from correctly completing
processing and receiving the actual information of the
That people can easily lose a target (the visual element
they want to find first in a sequence, process, then proceed
to the next) among masks (the surrounding elements, often
useful, but necessarily processed later in sequence) is a
confusion issue: we sometimes just do not know where to
look first or where to start sorting first in visual displays. But
this "confusion" is also illusory. For example, if the
arrangement of visual elements is such as leads the viewer to
resolve the whole set (the logical first target plus all nearby
masking elements) into one perception unit, that unit is an
illusion (often a powerful one) that blocks her from finding
the target. It is interesting that perception of a unit grouping
can supersede and block out search for an individual target.

Perceiving a pattern must be an extremely attractive and
compelling task.
Although completing patterns is attractive and often
useful, that it can also cause problems with illusion is
suggested by the experimental case in which people find it
difficult to find a C among 8's, 9s, and 6s (Cardosi, 1986).
The rounded shapes must tend to complete a pattern of
rounded elements, even though they belong to different
"artifactual (i.e., non-natural) categories" (letters versus
numerals) (Cardosi, 1986, p. 317). More space between
target and distractors or masks helps subjects to locate
targets more easily (Wolford and Chambers, 1983). That is,
more space should help to break up the pattern, lessen the
illusion that the whole set forms one perception unit, and
change the set back to individual elements, which the eye can
then more readily perceive and process individually.
Distance of viewing an element set may also influence
whether the set presents a confusing illusion to the viewer.
The "resting distance" (Raymond, 1986), the normal distance
at which a viewer regards a visual field, is the distance at
which visual accommodation is best. Problems with visual
accommodation to elements that are poorly spaced or that
tend to form a perception unit, and resulting blurring of
vision, may be one cause of perceiving misleading illusory
figures (boundaries or contours).

Visual material composed of a strong black and white
grating (alternating bar) pattern (Brown and Weisstein, 1988)
rises up out of the two-dimensional plane. If some elements
are stronger than others (gratings are more emphatic), the
stronger ones will appear closer to the viewer and produce an
illusion. This illusion masks information, because the sharply
distinct pattern calls attention to itself. Although its "third
dimension" carries no meaning, it SEEMS to, and it distracts
the viewer from information around it, and therefore slows
processing. The resting distance (the "best-accommodation,"
normal viewing distance; Raymond, 1986), the distance at
which the viewer is meant to view, no longer provides the
best perception, and perception becomes confused.
People processing information through perception
channels, whether operating under visual illusion or not, do
not merely process from the bottom up; we also process from
top down, that is, we are also influenced by memory (Wallach
and Slaughter, 1988). But as we process a visual field
containing a potentially illusory figure, if we can bring out of
memory the shape that the subjective contours induce, we
are more likely to perceive the subject contour "when the
'containing' pattern [is] shown" (Wallach and Slaughter, 1988,
p. 101).
Finally, subjects in trying to complete partial outline
symbols such as arcs of circles are distracted and make

mistakes (1) because of illusions as to where centers lie, and
(2) because of illusions of size caused by distractor circles or
circular arcs nearby (Roncato, 1983; Weintraub and Schneck,
1986). People look for edges and centers of pattern fields to
help them judge symbol centers and symbol sizes (Proffitt,
Thomas, and O'Brien, 1983), and they attempt to judge
relative sizes of area and volume (Butler and Overshiner,
1983). They are not very efficient in consistently judging any
of these correctly, again because of confusing illusions.
However, when people are directed to glance from one field
of symbols to another and identify which one element has
changed, results are correct, at far above chance level (Rayner
and Pollatsek, 1983). People are not distracted by an illusion
of a "set to be perceived as one unit" in this situation, because
they seem to be able to use a "memory buffer" to hold the first
arrangement in "view" just long enough to compare with the
second arrangement effectively.
Clearly, illusions are common in visual perception
situations. The confusion that they can cause varies a great
deal; sometimes its effect is negligible, as for example when
we complete incomplete outlines correctly and usefully many
times a day in our everyday activities. But when these visual
illusions appear in graphic material, in experimental
situations, they can slow processing. Because analogous visual
elements occur in geotechnical illustrations, and not singly

but in combinations of different element seems
unavoidable that illusions sometimes result and confuse
perception of such illustrations as well.
The Presence of Illusion in Geotechnical Illustration
Draftors of geotechnical illustrations are aware of the
possibilities for confusion, if not illusion, in preparing
illustrations. When mixing letters and symbols and numerals
in an illustration, for example, they leave as much space as
possible among the items, to reduce confusion and minimize
illusion caused by too many visual elements being read
together. They use serif as well as sans serif type, and roman
and italic type, to clarify and organize kinds of information
represented by "words"--italic for rivers and for mine names,
sans serif for structures, serif for towns, for example. In
other words, they vary the look of targets to minimize
distraction even within an alphanumeric category (letters as
opposed to numerals, as in Cardosi, 1986). Drafting a graph
using all sans serif type, and employing letters as well as
numbers as symbols is, as Cardosi would predict, hard to
read, because the targets are physically similar and each
becomes a distractor and elicits an illusion for the next. In
addition, arranging type along a line may ease recognition and
avoid illusion by framing the letters rather than leaving them

wandering among other letters (as suggested by research of
Bagnara, Simion, and Umilta, 1984).
Draftors commonly use symbols rather than letters in
illustrations, to locate points of interest; they use small solid
and open triangles, circles, and boxes, as well as plus signs,
circles with dot inside, and so on. If letters were so used,
and proliferated in an illustration, the reader would tiy to
encode the letter combinations as words, as suggested by
research of Johnson (1986), who noted that in a redundant
display of letters the subject tries to "encode the pattern as a
unit" (p. 94) before processing it. A related perception issue
in drafting is that a mountain range's name, for example,
must often be laid out across considerable space representing
that area on a map, and on some maps all capital letters are
used, while others use initial capital letter plus lower case.
From research such as Cardosi's (1986) and Johnson's
(1986), it can be concluded that it is easier for the reader to
achieve closure of a label using capital plus lower case type
when it is spread across an area than if all-cap type is used,
because capital letters with space between them look more
like symbols to be apprehended and interpreted one by one.
Moreover, if a draftor spaces the capital-letter type of such
words to make them lie across a large area, successive letters
are easier to dismiss as distractors of one another, similar but
successfully displaced from some as yet to be determined

"primary target" (as in Wolford and Chambers, 1983). And
therefore the reader's perception can miss the necessity of
processing these letters together as a unit.
Drafting of illustrations often employs many sizes of a
symbol, such as a circle, to discriminate classes of
information such as ranges of element values from sample
localities in a map area. Juxtaposition of circles often
misleads as to the size of any one circle, or overlap causes
distraction and confusion as to particular size-indicating-
value. From Weintraub and Schneck (1986), "context
supplied by the larger or smaller size of nearby nonconcentric
circles leads to size contrast" (p. 147); four or five sizes of
symbol circles in one illustration may induce considerable
illusion because of the interaction and inter-attraction of the
too-similar sizes at too close an interval.
A common error in author preparation of a map
illustration is to mis-scale the mapto prepare a line
representing scale that is scaled off incorrectly. Almost
always, the author proceeds to leave the scale wrong through
many iterations, though he or she, being extremely familiar
with the geographic area, should quickly realize that the
distance between points X and Y on the map, though actually
Z miles, scales out to Z+n miles. Such errors seem related to
people's difficulty with estimating areas represented in visual
materials (Butler and Overshiner, 1983), discerning edges

and centers (Proffitt, Thomas, and O'Brien, 1983), and
judging relative sizes of symbols to each other (Roncato,
1983), in this case a line representing scale to a known map
distance. Authors do not necessarily correlate a scale line
symbol correctly with a corresponding map area, a complex
"symbol" for a geographic area. There is no compelling
perception correlation of the two.
Several research articles, though they did not treat
illusion directly in their experiments, suggest how visual
materials can create illusion in illustrations; the illusions
insert their own meaning and distract the reader, blocking
her from processing the real information and apprehending
the intended meaning. Zucker, Stevens, and Sander (1983)
for example selected a moire pattern as one of their visual
elements in testing how subjects begin to perceive a series of
dots. Moire patterns in geotechnical illustrations occur
unintentionally when a piece of dotted pattern material is
overlaid at a certain inappropriate angle to another piece,
producing a rounded, very bright and confused pattern blotch
somewhat like the bright spots produced by the weave in
moire taffeta. Moires in geotechnical illustration patterns are
very noticeable and eye catching, and reviewers immediately
mark them for change. As map readers, editors and authors
reject them ("rotate pattern and eliminate moire!") because

moires' demanding illusions insert a distracting pseudo-
information which has no meaning.
People's ability to complete images from incomplete
stimuli (Uttal and others, 1988) definitely helps in
processing geotechnical illustrations containing many layers
of information, illustrations in which much basic information
has been screened so that more total layers of information
can be overlaid, discriminated, integrated, and understood.
Our ability to complete images suggests how we can quickly
process an illustration with a screened base, and links the
completion of masked stimuli into a usable image with the
successful perception of dots. We perceive where the dots
are (the screened base) and define their importance, and
then we perform a closure on the screened material, that is,
we accept the screened material as complete and as real and
informative, though ancillary. We process the screened
material along with or even before the main information,
saying in effect, 'The screened base is our context and
baseline; now what is new, what is to be learned?"
This "completing ability" (Uttal and others, 1988),
however, can also hinder us in reading illustrations. In some
map illustrations in geochemistry, for example, element
concentrations are shown by massed groups of small symbols
that can complete a compelling, distracting, illusory figure
such as a "dog's head" or an "umbrella." In such cases,

completing a pattern where the completion provides no
meaning in the context slows processing and creates
misunderstanding. Another unwanted sense of completeness
can occur when, in drafting, one pattern is applied across
several unrelated map units, or when one single color is used
for several unrelated units. We attempt to perform a closure
or see an identity where none exists.
Subjective contours and illusory boundaries (Petiy and
others, 1983) pose a considerable problem for preparing
coherent geotechnical illustrations. Not only can subjective
contours bar a reader from understanding a map illustration
replete with symbols showing geochemical concentration of
one element in potential mineral resource areas; also, when
illustrations must mix dense concentrations of different
symbols in adjoining or overlapping areas, subjective contours
may proliferate. If these "different" symbols are not "different
enough" (for example if one is a small triangle and one a small
diamond), the tendency is to perceive them ALL as the same
and to mistakenly achieve closure of an area (perceive a
subjective contour) that actually encompasses several areas of
meaning in the context. The reader might perceive the
concentrations of mercury and thallium as one meaning area.
Editors and draftspersons do well to scan illustrations for
subjective contours, especially to look for shapes that would
trigger a reader's memory for known objects (Wallach and

Slaughter, 1988), imposing a top-down memory bias on the
material that slows and confuses bottom-up processing.
Finally, some very small symbols used to draft an illustration
are difficult to resolve even if focused upon carefully, even if
presented to the fovea where "global and local processing can
occur in parallel" (Lamb and Robertson, 1988, p. 173). Such
tiny symbols blur into each other and whatever background
exists. Not only can they create an illusory figure that the
memory identifies and attempts to impose on processing, but
also, even that figure frustrates the reader because its local
elements are nearly impossible to sort, even with
considerable attention.
Illusory movement of stationary lines (Jiao and others,
1984) and illusory black and white grating patterns (Brown
and Weisstein, 1988) are not uncommon in drafted
illustrations. Illustrations in which several patterns
composed of distinct lines (in one pattern strongly vertical,
in another horizontal, in another at an angle) are illusory,
because as the vision is led in several directions at once, each
direction seems to provide emphatic linear information
which in fact has no meaning. Using one or more strong
grating patterns for units on a map is illusory because the
grating patterned areas seem closer to the reader than
others, whereas the patterned areas should be on one and the
same plane for clear processing. The sharply distinct

8 1
patterns claim attention from information around them and
give the illusion of greater importance.
One particular pattern that draftors sometimes apply to
a map unit produces an instant illusion; it is a screened
pattern of small broken circles, with adjacent circles broken
in different places. In reviewing a drafted geologic map, I as
editor always reject this pattern, because an illusion of letters
(the rounded onesC's, D's, G's, O's, P's, Q's, U's) within the
pattern appears, promising some sort of meaning, however
simple. With part of my processing capacity looking for
letters to complete and recognize, I miss the over-all view of
what the pattern doeswhich is merely to cover the area.
The broken circles simply are the elements of the pattern,
nothing more. This illusion seems to reflect a situation of
visual accommodation being inaccurate "when stimuli were
viewed at a distance that corresponded to [other than] the
individual's resting position" (Raymond, 1986, p. 281). I
strain to read the pattern's meaning and there is no meaning
there to read. On the other hand, large, bold type is
occasionally selected and applied to maps (usually done
inadvertently by apprentice draftors unsure of the final scale
of the illustration). Such type looks odd and intrusive, as do
thick black arrows used to indicate features of interest.
These type and symbol elements are always marked for
exchange to "standard" type sizes and line weights. Standard

sizes of type, weight of lines, and width and density of
patterns must have been intuitively chosen by draftors
generations ago, to be readable and clear at a reader's resting
positionand not at a viewing distance of 2 inches or 15 feet.
The reality of quick, automatic "eye movement yet
perception of continuity" investigated by Rayner and Pollatsek
(1983) may suggest how we perceive adjoining pattern areas
in illustrations. Readers sometimes are confused when they
glance from one pattern to another adjoining one, because
the patterns are "too much alike" and induce an illusion of
identity, or because one is distracting in some way (such as by
the illusory "alphabet soup pattern" just mentioned, or a too-
glaring or too-bland color-on-color pattern). A determinable
level of pattern discrimination must exist that would
minimize confusion in the reader's memory buffer (Rayner
and Pollatsek, 1983) during eye jumps from one area and
As illusions in perception become noticeable and real in
psychophysical experiment, so they do in geotechnical
illustration. Despite considerable drafting attention to avoid
misleading the readers, illusions can and do appear. And that
they do appear seems explainable by the propensity with
which people automatically perform many complex tasks in
processing visual informationsearching for literal and
figurative centers, identifying radii, and completing

incomplete images. The more complex an illustration is, the
more layers of information and the more kinds of visual
material it must integrate, the more chances probably occur
that we will perceive illusions and confuse ourselves. The
psychophysical research makes the situations where illusions
are prone to appear more obvious, and therefore indicates
how editors and draftors can avoid them more effectivelyby
watching for those problem situations.

Research articles in psychophysics and human
perception address one issue at a time: they experiment with
one visual element or a small closely related range of
elements at a time. Geotechnical illustrations use multiple
types of visual materials to convey information. Thus, the
relation of the research testing to illustration is not a matter
of considering a certain number of simple relationships taken
separately, but a matter of integrating those relationships to
describe more fully what takes place in perceiving a complex
geotechnical illustration.
The more complicated visual elements that are used in
psychophysical testingthe more strokes used and the more
they proliferate across a field, and the closer they come to
containing meaning in themselvesthe more the
psychophysical research becomes relevant to more complex
issues in geotechnical illustration, particularly those of clarity
and coherence in multi-layered illustrations. (A clear
illustration will be one in which the reader can separate all
levels of information, grasp what information is represented
at each level of emphasis, and find that all information of the

same kind is represented analogously. A coherent illustration
is one in which the reader is able to integrate all kinds of
information with minimal processing effort, to receive the
overall message of the illustration.)
Moreover, research that exposes, for example,
perception problems that subjects have with patterns not
only comments on how patterns function in geotechnical
illustration but also is meaningful in combination with
research on the other visual elements, because all these
dots, lines, patterns, symbols and letters, and colorvery
commonly appear together in illustrations (and may be
accompanied by misleading illusion). Some individual
articles, and many of them considered together, are relevant
to the whole issue of the success or failure of layering
combinations of elements onto an illustration, to achieve
coherent three-dimensionality on a two-dimensional surface.
A parallel issue is the planning of series of related
illustrations, which adds a variable that I call "horizontal
layering," that is, the processes by which related illustrations
achieve coherence as a series. Each illustration of such a
series also must individually achieve internal layering
coherence through effective "vertical layering."

How Psychophysical Research Underlies Effective Layering in
Geotechnical Illustrations
Perhaps the most straightforward example of research
useful to understanding layering notes that, 'The ability to
perceive wholeness in a world of naturally masked stimuli and
noisy or incomplete views of objects is obviously a very
valuable skill" (Uttal and others, 1988, p. 223). In everyday
life, we constantly perceive wholeness and regularity, as well
as consistent color, in a welter of occluded, rotated,
inconsistently illuminated, and otherwise distorted images.
We also complete and integrate what we see when we look
straight ahead and also to the sides (Indow and Watanabe,
1984). We see a continuity and a wholeness when the
"reality" is neither continuous nor whole. Even our vision is
not precisely continuous, because the eye jumps we
constantly make from one area of our visual field to another
entail milliseconds of blurred vision (Rayner and Pollatsek,
We do pick areas to focus on first in a visual field.
Proffitt, Thomas, and O'Brien (1983) experimented with
subjects' perception of edges and centers (particularly
centers) of two-dimensional pattern fields; they began, "The
visual system does not equally apprehend all points within
bounded shapes, but, rather, certain locations are of special
importance for perception" (p. 63). The determination of

centers in experiments was much more closely tied to
perception of boundaries than to luminance: "Our
manipulation to vary the luminance distribution with shapes
did little to affect the location of perceived centers from their
configurally specified locations" (p. 70). Subjects looked for
centers according to shapes and "configurations."
As we select a center-of-emphasis point in a patterned
area, analogously in perceiving a layered illustration we in
some sense must select points within each layer to begin
processing each layer. Alternatively, we process one whole
layer first, which in any case is a selection of emphasis or area
for first processing in a sequence.
That people find it important to perceive centers is
expanded by work on their ability and tendency to perceive
depth. Braunstein and others (1986) showed subjects
rotating images on film and asked them to determine
direction of turn by "edge occlusion" in "opaque images" and
"element occlusion" in transparent images" (p. 216). They
According to Marr (1982 ["Vision," published by
W.H. Freeman]), a viewer-centered
representation, the '2 1/2-D sketch, is recovered
from information in the retinal image, and is used
to build an object-centered representation, the
'3-D model (p. 216).
I suggest that people thus process and understand depth cues
in visual material in a kind of continuum, processing from

images of objects to processing layers of information in a "2
1/2-D" illustration.
Although people readily attempt to perceive centers
and find focus points, and readily perceive outwards from
them, their judgment of area and volume are often faulty
(Butler and Overshiner, 1983). Their linear estimates are
reasonably good, but error escalates as mental multiplication
is done to find area and volume; seemingly they do not
automatically move accurately through visualization of two-
and three-dimensional visual material.
Subjects in experiments by Toye (1986) made
judgments on distances between stakes set in the ground in a
visual array outdoors. Some subjects judged these distances
from one position only and some from two positions set 90
degrees apart. As a long-time editor, I was pleased that Toye
asked his subjects to make maps of the array, because I
thereby found his work easier to extrapolate to geotechnical
illustration. Toye wanted to test the "classical psychophysical
model of space perception" against a "strong ecological
model"; his results "emphasize[d] the importance of the
subjective element of visual space perception" and supported
the "classical Helmholtzian psychophysics" that "reduces
visual space to a Euclidean coordinate system" (p. 85).
Viewing scenes at 90 degree differences, Toye believed,
would test classical psychophysics against the ecological

model because the ecological model "relies on invariant
affordances common to all scenes" (p. 86) whereas in
classical psychophysics depth and breadth are "perceived
differently" and a different viewpoint should change "involved
psychophysical mechanisms" (p. 86). Toye noted that in the
classical model, "Depth information is provided by hard-wired
mechanisms such as stereopsis and convergence ([references
here]) and local cues such as motion parallax, overlap, and
perspective" (p. 85). The maps the subjects made did vary in
accordance with their viewpoint, which suggests that depth
cues are important for visual processing but that they are not
invariant and that they vary with viewpoint. I suggest that
how we use them also varies depending on where in layers of
a multi-layered visual field we decide to process. Being able
to sort depth cues quickly and consistently speeds
Overload and confusion can occur in perception if
masks or distractors slow subjects' location of targets.
Subjects of Wolford and Chambers' work (1983) had less
trouble locating a target among masking elements if space
between the target and masking item was increased. Two
nearby items "are presumed to compete for a limited set of
feature detectors" on the retina (p. 129). At close spacing of
target and mask, this "feature interaction" may dominate our
ability to group data. If targets of varying importance occur in