than“)? a...
fi.'l. I
i ...$..w¥an

. ix .1 Vans .
...&..«h.‘¢.h«\‘r.kn, 4
“i” in"? fis

I. 2“»
3. V2.7
inWJ‘MWvgi.
t

5%,;

zirt
,o

3 ‘
.lx.....u.r..-fl_. I...
-. «nut/3.14..
. _ €104.23. an“? n .
2.64:; 0.1.3.1. ....
3.». . a. 1a.

Li‘ .1: if . “I J.
In IL..E! n< .
1 4.1.1:. I. I u “H
. . 5| . (t4? 91:81..
. .. fi‘nvat(.¢§’1vn‘ .(
~. 4.; hum». » 1...“.
Soaihws 3:
.17. . slut .
L) to!)
A (Big-1
. if5
:4“? N

.15“
5 In; . '11 ‘ fin
s. I. :5...- . A
.Oiizll..." .

killing...
:ull

‘1. i I

 

_ 1.1m? firuimu.)

it .

2. =9. have ifwfifa:

. anrvyar in?! ._ t. ..K

 

 

 

m. . LIBRARY
mi“ Michigan State
20 0 <5 ‘ University

This is to certify that the
dissertation entitled

WRITING MEMORY: A STUDY OF MEMORY TOOLS
IN INVENTION

presented by
STEWART NEAL WHl‘l‘l'EMORE

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Rhetoric and Writing

 

 

Germ/0‘

 

 

{LU Nfajorl/Professor’ 5 Signature»
\« M Fl 2: cl?

Date

MSU is an affinnative-action, equal-opportunity employer

'0- -0-‘---o-o-0_l--|-I--A

a..- -.—-.-.--.—.-...

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 Kthroj/Acc8PreleIRCIDateDuetindd

WRITING MEMORY: A STUDY OF MEMORY TOOLS IN INVENTION
By

Stewart Neal Whittemore

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Rhetoric and Writing

2008

ABSTRACT
WRITING MEMORY: A STUDY OF MEMORY TOOLS IN INVENTION
. By
Stewart Neal Whittemore
There is a growing recognition that one of the key limitations for all writing-

based knowledge work, whether in schools, workplaces, or communities, is how to
productively handle information overload imposed by information technologies as
research and writing tools. Rather than conceptualizing information load in terms of
data storage — a move that severely curtails a focus on memory in contemporary study
of rhetoric — this study builds on the idea that stored information or memories are only
useful when they are employed in support of rhetorical practices—that is, when they
can be retrieved and used in specific situations to solve problems and meet audience
needs. Therefore, I reason that a productive way to research memory-in-use is by
paying close attention to specific 'scenes' of memory work in which writers retrieve and
use stored knowledges in rhetorical situations using all the affordances perceptible to
them in their embodied contexts, much as ancient orators used spatial structure as a
mnemonic to recall the points of a speech. To accomplish this, I studied the memory
practices of a team of technical communicators in a medium-sized software firm over a
six-month period. The data from this research, including recorded observations and
interviews supplemented by collected artifacts and field journals, were analyzed to
identify the role of embodied contexts in writers’ memory work in five scenes of
composing. The results of this analysis contribute a new theoretical and methodological

foundation for studying memory work as rhetorical practice.

Copyright by
STEWART NEAL WHITTEMORE
2008

To
Patrick Joseph Canney
1940 - 2007

Bonae Memoriae

ACKNOWLEDGMENTS

I would like to thank my research participants who welcomed me as a member of their
team and gave me so much time and support — you are the most talented technical
communicators I have ever met and an inspiration. I would like to thank my family,
Erin, Owen, Patsy, and my mother and father, for their unflagging support and
encouragement through four hard but rewarding years. I would like to thank my
committee members, Jim Porter, Bill Hart—Davidson, and Julie Lindquist, for their
advice and friendship. Lastly, I would like to thank my committee chair, Jeff Grabill,
without whose praise, encouragement, and brilliant example I would never have begun

this odyssey.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ viii
LIST OF FIGURES ......................................................................................................... ix
CHAPTER 1: INTRODUCTION AND REVIEW OF LITERATURE ........................... 1
Memory Work and Technical Communication ............................................................ 1
A Theory of Memory in Invention ............................................................................... 5
Literature Review of the Problem of Information Overload and Invention ............... 11
The Need for a Better Understanding of Memory Work as Rhetorical Practice ........ 18
CHAPTER 2: THEORETICAL FOUNDATIONS ........................................................ 21
Dimensions of a Rhetorical Practice of Memory ....................................................... 21
The Role of the Body in Writing and Memory .......................................................... 24
Aristotle’s Embodied Epistemology ........................................................................... 31
The Aristotelian Phantasm and Social Theories of Cognition ................................... 35
CHAPTER 3: METHODOLOGY AND RESEARCH METHODS .............................. 43
Scenes of Memory Work ............................................................................................ 43
The Research Site ....................................................................................................... 50
The Research Participants .......................................................................................... 55
Physical and Digital Infrastructures ........................................................................... 56
Data Collection Methods and Rationale ..................................................................... 59
Data Collection Schedule ........................................................................................... 66
Data Preparation ......................................................................................................... 67
Data Segmentation and Coding .................................................................................. 68
Data Analysis .............................................................................................................. 75
CHAPTER 4: TWO SCENES OF MEMORY WORK WITH LANCE ........................ 77
Introduction ................................................................................................................ 77
Scene 1: Arguing for Changes to the User Profiles .................................................... 80
Scene 2: Meeting with Wallace .................................................................................. 87
Analysis of Lance’s Memory work ............................................................................ 93
Conclusion ................................................................................................................ 109
CHAPTER 5: TWO SCENES OF MEMORY WORK WITH LUCY ........................ 111
Introduction .............................................................................................................. 1 1 1
Scene 1: Updating the User Guide ........................................................................... 118
Scene 2: Attempting to Solve a Problem .................................................................. 122
Analysis of Lucy’s Memory Work ........................................................................... 125
Conclusion ................................................................................................................ 143
CHAPTER 6: CONCLUSION ..................................................................................... 145
Introduction .............................................................................................................. 145

vi

What the Study Revealed about an Embodied Rhetoric of Memory ....................... 146

What the Study Revealed about the Memory Regime: A Final Scene ..................... 149
Analysis of the Scene ............................................................................................... 154
Implications for Practicing and Teaching Technical Communication ..................... 161
APPENDIX: INITIAL INTERVIEW QUESTIONS ................................................... 164
Questions about Participant Background ................................................................. 164
Questions about Composing and Writing Practices ................................................. 164
Questions about Project Management Practices ....................................................... 165
REFERENCES ............................................................................................................. 167

vii

LIST OF TABLES

Table 1: Research Participants ....................................................................................... 56
Table 2: Data Collection Methods Mapped to Scenes and Research Questions ............ 61
Table 3: Data Collection Sessions .................................................................................. 67

viii

LIST OF FIGURES

Figure 1: The scene of memory work ............................................................................. 47
Figure 2: Information development wing ....................................................................... 57
Figure 3: First part of scene one - Lance’s office .......................................................... 81
Figure 4: Second part of scene one — Scrum hall ........................................................... 81
Figure 5: New section of the user guide ......................................................................... 83
Figure 6: Lance’s reminder document ............................................................................ 85
Figure 7: Annotations to Lance’s reminder document ................................................... 86
Figure 8: Wallace’s office .............................................................................................. 88
Figure 9: Lucy’s office ................................................................................................. 118
Figure 10: Lucy “draws” the Sprint folder ................................................................... 139
Figure 11: Lucy “draws” the Features folder ............................................................... 139
Figure 12: Lucy “draws” the subfolders ....................................................................... 139
Figure 13: Left-handed gesture .................................................................................... 140
Figure 14: Two views of Lucy’s data dump folder structure ....................................... 141
Figure 15: Angela and Carl’s meeting in the petri dish ................................................ 152

ix

CHAPTER 1: INTRODUCTION AND REVIEW OF LITERATURE

Memory Work and Technical Communication

Arguably, the most important sector of labor in an information economy is work
that centers on the storage and maintenance of knowledge and its timely and effective
retrieval for new purposes. Inarguably, writing is the prototypical and most effective of
all technologies for preserving and retrieving information and knowledge: in other
words, for doing memory work. As Ong (1982) puts it: “by taking conservative
functions on itself, the text frees the mind of conservative tasks, that is, of its memory
work, and thus enables the mind to turn itself to new speculation” (p. 41).
Consequently, writers within contemporary organizations play a vital role in
maintaining and using the memories of those organizations. This is especially true of
technical communicators, who perform for their organizations the critical work of
finding, assessing, filtering, and translating information stored in texts in various modes
(e.g., technical specifications, user requirements, task analyses, workflows, fields in
databases, etc.) into useful, actionable knowledge for diverse audiences in contingent
and variable situations. In fact, earlier traditions of workplace writing saw this work of
retrieving and adapting existing knowledge to the exigencies of shifting rhetorical
situations as so important that it was given its own canon in rhetorical theory, the canon
of memory, and skill in memory work was considered absolutely essential to invention,
the creative process by which orators or writers determined what to say and how to say

it to meet the needs of their audiences.

Yet, despite the importance of such creative work to contemporary

organizations, the memory work that writers perform is often not recognized or

rewarded as critical to an organization’s success. This is an unhealthy development for
both writers and organizations. It is unhealthy for writers because the memory work
that writers, particularly technical communicators, perform is at the core of what
distinguishes writing as symbolic analytic work, the highest form of labor in the global
information economy according to former Secretary of Labor Robert Reich. Reich
(1991) draws a direct link between the activities of symbolic analysis and the memory
work of writing by describing symbolic analysis as work that entails the timely,
effective, and creative reuse of existing information for new purposes: “Discovering
patterns and meanings [in stored information] is, of course, the very essence of
symbolic analysis” (p. 229). Because symbolic analytic workers are the highest in
status, pay, and prestige of the three broad strata of labor that Reich discerns in an
information economy, performing (or being perceived as performing) this memory
work is critical to the long-term status, pay, and prestige of technical communicators in
their workplaces. Correspondingly, failure to recognize or understand the memory
practices by which writers retrieve, assess, and repurpose information into timely and
useful knowledge products is unproductive for organizations because it prevents those
organizations from fully benefiting from the inventional expertise of one of the most
important sectors of labor in an information economy. So, making the memory
practices and tools of writers visible should be a critical task of any research focusing

on technical communicators.

Such research can borrow from long and varied research traditions exploring
writing, memory, and inventional practices in workplaces. First, there are workplace

studies that combine ethnographic methods with various types of textual analysis to

study the ways in which people construct and preserve knowledge in organizations
through writing and communication. Studies of this type include Zuboff’s (1988) now
classic study of the effects of automation and information technology on workplace
knowledge practices and Sellen and Harper’s (2002) exploration of the importance of
the materiality of paper documentation to knowledge work. Similarly, Winsor (2000,
2003) and Spinuzzi (2003a, 2003b) explore the role of workplace genres in preserving
organizational knowledge and power relations while Brown and Duguid (2002)
examine the ways in which knowledge is created, preserved, and transmitted through

networks of communities of practice.

Next, there are historical and archival studies that analyze texts from the distant
and not-so-distant past in order to trace the development and evolution of memory
practices in various types of workplaces from various epochs. Although I may be the
first person to ever label her scholarship “workplace research,” Mary Carruthers’ (1990,
1998) exploration of the memory practices of Greek and Roman orators and medieval
monks provides fascinating insight into the tools and practices of knowledge
management in the ancient workplaces of forum, pulpit, and scriptorium. Carruthers’
central finding is that the knowledge management issues of storage and retrieval,
though often thought of primarily as a byproduct of the computer age, are in fact not
new at all but have been concerns for effective workplace practice since the beginning.
More conventional of archival-based workplace research is that of Joanne Yates’ (1989)
who studied the rise of managerial command and control during the late nineteenth and
early twentieth centuries via such mundane yet crucial technologies as vertical filing

cabinets and internal genres like the work order. More recently, Bowker (2005)

explores the memory practices of scientific researchers ranging from nineteenth century
geologists to twenty-first century environmental scientists through analysis of the

scientific report genre.

Finally, there is the strand of research that attempts, through close observation of
writers at work, to model and describe the cognitive processes underlying the writing
process, including the influence of a writer’s pre-existing knowledge or memories on
invention. Flower & Hayes (1981) inaugurated the golden age for this type of research
by recording writers thinking aloud during composing sessions conducted in lab-like
controlled situations. Hayes (1996) and Kellogg (1996) continue this tradition,
attempting to further illuminate the role of constructs such as affect and short-term
memory during composing. Other researchers have attempted similarly close
observations of writers under more natural conditions: in situ in the writers’ actual
workplaces and by substituting methods other than unnatural think-aloud protocols. Of
particular interest for a study of the role of memory in composing are the work of Haas
(1996), who describes aspects of writers’ embodied interactions with computer
technologies of writing that are habitual, or involving bodily memory in some form, and
the work of Prior and Shipka (2003) who study the ways in which composing and
invention are distributed across space and time, an approach that has important

implications for understanding the functioning of memory.

However, despite the richness of these research studies exploring the
interrelationship of writing, invention, and memory there have been few attempts to
bring the strands of research together in order to explore the tools and practices by

which contemporary workplace writers, like technical communicators in knowledge-

intensive organizations, manage the complex “datacloud” of “shifting and only slightly
contingently structured information space[s]” of the contemporary workplace in order to
perform the creative work of symbolic analysis — to invent new things by writing with
and about older things (Johnson-Eilola, 2005, p. 4). As a consequence of this, technical
communicators’ contributions to the memory work of organizations risks, at best, going
unnoticed, or, at worst, being viewed as low-skill non-creative work rather than the

dynamic and creative form of labor that Reich conceptualizes as symbolic analysis.

A Theory of Memory in Invention

Understanding workplace writing as memory work first entails articulating a
theory of invention for composing in contemporary knowledge-intensive workplaces.
Reich’s (1991) delineation of the four characteristics of symbolic analytic work offers a
good starting point for this endeavor. According to Reich, symbolic analytic work is
first characterized by the twin moves of abstraction and system thinking,
complementary processes by which raw information is first analyzed in order to find
patterns and hidden meanings and then communicated and synthesized within larger,
interconnected systems in order to discern potential areas for innovation. Symbolic
analysis also entails experimentation, in which established bodies of knowledge are
tested and played with “in order to better understand causes and consequences” (Reich,
1991, p. 232). Finally, symbolic analysis requires various forms of collaboration, in
which information that would otherwise be overwhelming to a single person working
unaided, is divided up and dispersed among other tools and actors before being
reassembled into useful knowledge. Johndan Johnson-Eilola (1996) makes the

connection between Reich’s quadripartite formulation of symbolic analytic skills and

the invention processes of technical communicators explicit: “the ability to manipulate,
abstract, revise, and rearrange information is itself one version of the classic task of the
technical communicator: someone who takes pre-existing knowledge about technology

and explains it to others” (p. 255).

Reich’s emphasis in this theory of invention is not that stored information — here
conceptualized as memories — has become less important to creativity in the information
economy — indeed the ubiquity of access to stored information in contemporary
workplaces actually makes it more important — but rather that stored information must
be rendered usable and useful. And, according to Reich (1991), it is the job and the
distinctive skill of the symbolic analyst to do this work of making memories usable:
“The symbolic analyst wields equations, formulae, analogies, models, constructs,
categories, and metaphors in order to create possibilities for reinterpreting and then
rearranging, the chaos of data that are already swirling around us” (p. 229). What
Reich’s theory of invention does not give much sense of, however, are the specific
practices by which symbolic analysts perform this work of rendering stored information
usable in specific times, circumstances, and situations to meet the needs of audiences.
In other words, what Reich’s theory lacks is an understanding of such memory work as

rhetorical practice. This is what turning to the rhetorical tradition can give us.

Rhetorical practices for making memory usable in invention are hardly new -
each age develops its own set of practices to manage the information made available by
new technologies of storage and retrieval in order to meet the needs of new audiences
and new situations. Jocelyn Penny Small (1997) summarizes this phenomena

succinctly: “each tool or technology makes it possible to deal more efficiently with the

current accumulation of words, but by virtue of its success propagates yet more words
that need yet more techniques to control them” (p. 141). Reich’s articulation of
symbolic analytic work, therefore, is enriched and expanded when considered in
relation to earlier rhetorical practices that make more explicit the connection between
the retrieval and use of stored information for audiences as situated acts of invention.
Of special relevance to this are the rhetorical theories and practices of the canon of
memory, one of the five essential divisions of classical rhetoric, particularly as it was
developed in Roman rhetorical theory in response to the specific exigencies of the
occasions of Roman civic oratory. Catherine Steel (2006) describes the exigency of

Roman oratorical situations in this way:

A speech is prepared for a specific time and place, to be directed at a specific
audience and [. . .] with the aim of securing a specific outcome. Moreover, this
first performance is, logically, oral and does not imply the existence of a written
text; indeed, there was a strong convention within ancient rhetoric that speeches
were delivered from memory, and even though written texts might well feature
in preparation, orators would often find themselves in situations where

improvisation was necessary. (p. 25)

This ensured that Roman oratory developed around the idea that merely possessing a
reserve of stored content, however vast, was not adequate in and of itself: that
knowledge was only useful when it could be retrieved, adapted, and used at the precise
moment in which it was needed in order to meet the needs of the particular scene of
rhetorical activity. This moment was the moment of kairos, which, as Kinneavy (1986,

2002) points out, was a central concept in Roman rhetoric. Kinneavy (1986) interprets

kairos as a two—part concept in Roman thought emphasizing both time and
appropriateness: “the right or opportune time to do something, or right measure in doing
something” (p. 80). Time and appropriateness together, Kinneavy (1986) points out,
delineate the boundaries of a rhetorical situation: “the appropriateness of the discourse
to the particular circumstances of the time, place, speaker, and audience involved.” (p.

34).

While at first glance, kairos as a concept would seem to be more significant to
the canon of delivery than those of memory or invention, recent rhetorical scholarship
has located Roman thinking about the creative processes of invention much closer to
contemporary social constructivist theories than to post-Enlightenment notions of the
isolated creative and unified individual mind. That is, similar to social constructivist
thought, which holds that invention is a process “initiated by inventors and brought to
completion by an audience,” implicit in the Roman view of invention was the notion
that speaker and audience together create new things (LeFevre, 1987, p. 63). First,
classical theories of invention started with the notion that writers and orators “began
their investigations with what other people thought, rather than with an introspective
review of their own thought processes” (Crowley, 1990, p. 16). The purpose of the
speech, therefore, would be to “jog the memories of both rhetor and audience, since it
would mirror the way ideas had been stored there in the first place. Memory was
especially well served when the rhetor employed the orders of place or time” (Crowley,
1990, p. 44). Similarly, Carruthers (1990) describes the process of invention during
composing as a joint process rather than as a soliloquy occurring in the orator’s head:

“for composition is not an act of writing, it is rumination, cogitation, dictation, a

listening and a dialogue, a ‘gathering’ (collectio) of voices from their several places in
memory” (p. 198). In a sense, then, mnemonics of the ars memoria, such as the
place/image or walking mnemonic, were tools for invention as much as they were tools
for storage, as speakers and audiences together walked their memory places creating

shared meanings and understandings.

So, despite recent misinterpretations of the rhetorical tradition, such as Yates
(1974) or Corbett & Connors (1999), who characterize the techniques of the ars
memoria as technologically outdated and overly complex methods of storing large
quantities of information, in its classical origins the rhetorical canon of memory was in
fact primarily concerned with retrieving stored information for kairotic purposes in
order to facilitate invention during rhetorical situations. As Carruthers (1990) puts it, in
classical and medieval rhetorical theory “the proof of good memory lies not in the
simple retention even of large amounts of material; rather, it is the ability to move it
about instantly, directly, and securely that is admired” (p. 19). Similarly, Crowley and
Hawhee (2004) note that classical rhetoric emphasized a “memory-ready” condition,
not unlike the ready stance [of] kairos” (p. 317). In this interpretation, therefore, it is
kairos that is most important in the memory work of invention processes because, to be
useful in invention, information must be retrieved and transformed in the situational and

contextual moment.

Similar to contemporary activity theory, the classical rhetorical tradition also
emphasized that memory work was shared and distributed: not just between speakers
and audiences but also between writers and their work spaces and tools, and between

writers and their bodies. Cicero, for instance, emphasizes the somatic and spatial nature

of human memory — the way in which information is best understood when it is
connected to the world of the senses, of artifacts and spaces. Cicero (1942) states “the
most complete pictures are formed in our minds of the things that have been conveyed
to them [...] by the senses , [. . .] But these forms and bodies, like all the things that come
under our view require an abode, inasmuch as a material object without a locality is
inconceivable” (p. 469). The entire basis of the walking mnemonic was the ancients’
theory that “the whole sensing process, from initial reception by a sense-organ to
awareness of, response to, and memory of it, is somatic or bodily in nature” (Carruthers,
1990, p. 48). Small (1997) adds nuance to this: “Once the places take on a physical
form, they become subject to the physical limits of human perception, for the [author of
the Rhetorica ad Herennium] believes that the external and the internal eye match in

their abilities” (p. 100).

Finally, classical rhetoric emphasized that information had to be personalized in
some way in order for it to be useful to anyone, speaker or audience. Carruthers (1990)
makes this point best when she illustrates that, for the ancients, “memoria refers not to
how something is communicated, but to what happens once one has received it, to the
interactive process of familiarizing —- or textualizing — which occurs between oneself
and others’ words in memory” (p. 13). This textualizing process is similar to the process
of abstraction in symbolic analytic work in that information “out there” must be
internalized in order to become susceptible to manipulation, to use. Reich (1991) puts it
this way: “The real world is nothing but a vast jumble of noises, shapes, colors, smells,

and textures—essentially meaningless until the human mind imposes some order upon

10

them. [. . .] Reality must be simplified so that it can be understood and manipulated in

new ways” (p. 229).

Together, then, these theories of invention, ancient and contemporary, suggest a
conceptual framework for evaluating memory work as a rhetorical practice of invention.
This framework consists of three basic insights. First, it suggests that, in memory work,
the responsiveness of stored information to retrieval and manipulation in actual
situations of use is more important to invention than raw storage. Phrased more
succinctly, the interface is more important than the database for performing memory
tasks. Second, it suggests that memory is most useful when we can offload it in some
way to our surrounding material and social environment. We offload memory to our
material environment through our use of our tools and workspaces; we offload to our
social environment by sharing and communicating memory with the audience of our
peers in order to share the labor of walking our memory places. And finally, it suggests
that memory is inherently psychosomatic, involving both the mind and the body, and,
therefore, that information must be susceptible to personalization in some form, that
there must be a way to “touch” and to “feel” data in the world of the senses. Any tool,
workspace, infrastructure, or job description that does not conform to these three

insights will not be entirely successful in supporting memory work.

Literature Review of the Problem of Information Overload and Invention

The principal challenge facing the symbolic analyst is a problem of memory:
how to manage the information load of a global business environment characterized by
the “explosion of networked information and affordable computational and storage

resources” without becoming overloaded (Larsen & Wactlar, 2003, p. 8). In such an

11

environment, so-called “information overload” is an ever-present threat — as well as an
opportunity for the symbolic analyst to display her most valuable skill. All three
conditions for memorial invention are affected by information overload: if we are
overloaded with information, we cannot respond to rhetorical situations in a timely
fashion; if we are overloaded with information, offloading for purpose of “filtering and
assessing” this information becomes difficult if not impossible; and finally if we are
overloaded with information, we can no longer grasp data in the world of the senses in

order to see the big picture (Gee, Hull, and Lankshear, 1996, p. 38).

Computer technology provides, of course, both the font of access to information
as well as the geyser that leads to information overload in contemporary workplaces.
The literature on contemporary digital interfaces for memory work sheds light on the
phenomena of information overload and reveals several disciplines trying to formulate
better digital workspaces for performing memory work. In particular, two fields have
contributed substantial insights into the design of interfaces for managing memories
digitally: library and archival science and the somewhat nebulously defined field of
“office automation,” centered on the various special interest groups (SIGs) of the
Association for Computing Machinery (ACM), including especially Information
Retrieval (SIGIR), Computer-Human Interaction (SIGCHI), and those contributing to

the Conference on Computer-Supported Cooperative Work (CSCW).

The field of library and archival science has, it goes without saying, witnessed a
sea change over the past fifteen years with the rise of the Internet as the
overwhelmingly dominant method for gaining access to and utilizing the resources of

archives. In the last few years, studies have begun to appear in the archival science

12

journals that examine archival interfaces from a variety of perspectives, including
studies analyzing political aspects of interface design, as collections which were
formerly heavily-curated and closely-managed are made available to anyone with an
Internet connection (Delmas, 2001; Besser, 2002; Hedstrom, 2002; Bizjak, 2000),
critiques or refinements of specific interface elements for finding information, such as
augmentations or replacements for the ubiquitous search and browse paradigm
(Coleman & Oxnam, 2002; De Chiara & Scarano, 2004; Fast & Sedig, 2005; Lansdale,
2005; Matusiak, 2006); and direct user-research studies such as focus groups and
usability tests of archival interfaces (Kani-Zabihi, Ghinea, & Chen, 2006; Glosiene &

Manzhukh, 2005; Beaudouin-Lafon, 2004; Hong, Thong, Wong, & Tam, 2002).

Each of these studies identifies significant problems with the design of current
interfaces for finding and managing information, but perhaps the most complete
summarization of the interface-related problems facing information and archival science
appear in Larsen and Wactlar’s (2003) report of the NSF Workshop on Research
Directions for Digital Libraries, “Knowledge Lost in Information,” which summarizes
the difficulties facing users by the “explosion of networked information and affordable
computational and storage resources” and articulates a research agenda to guide the
development of the next generation of archival interfaces (p. 8). Larsen and Wactlar
note that simply dumping data into Internet-accessible digital archives is not adequate;
that users need to be provided with tools for finding the appropriate data at the
appropriate time and for evaluating and using that data to support meaningful activities.
They add, “data of many types will be increasingly abundant and ‘technologically’

available. But these data will continue to seem chaotic, lacking sufficient organization,

13

stability, and quality control. Moreover, individuals and communities may lose the
ability to control access to and manage their own data” (p. 8). In response to these
challenges, Larsen and Wactlar (2003) articulate a five point research agenda, including
exploring methods of “employing ‘context’ in information retrieval at the technical,
individual, and societal levels, [. . .] integrating information spaces into everyday life,

[and] reducing data to ‘actionable’ information” (p. 9).

Larsen and Wactlar’s (2003) recommendations reveal a new recognition among
archivists that the most important issues facing contemporary archival science lie not
with raw storage but at the interface where users interact with the “information ether”

(p. 1). Significantly, Larsen and Wactlar (2003) conclude by arguing:

The volume, complexity, and heterogeneity of new information outpaces even
the most advanced of current approaches. Part of the solution may lie in better
tools for envisioning information spaces. Research in this area posits that search
can be improved by exploring new ways (or old ways in new data environments)
to visualize media-rich information. [. . .] replacing information overload with an

intuitively understandable visualization that captures the essence of a situation.

(P- 9)

This need to capture the essence of a situation leads Larsen and Wactlar (2003) to make
an interface suggestion of their own, one that, as will be shown, seems to have

interesting parallels with the techniques of the ars memoria:

We know that information visualization (in the full multimedia sense) can be

effectively employed to summarize content and provide the means to display it

14

in new and novel ways. Imagine an information room in which users can
specify an initial domain [. . .] and ‘walk through’ the information space. [. . .]
and, perhaps, discover additional information that bears on their interests as seen

through the prism of the visualized or sensed world. (p. 15)

The ACM’s work on digital office automation also reveals parallels with the
rhetorical tradition. In particular, one of the foundational dialogs in the ACM about the
function of digital memory interfaces stems from Malone’s (1983) study “How Do
People Organize their Desks? Implications for the Design of Office Information
Systems.” Malone notes that one of the most insightful and surprising findings of his
study was the degree to which the spatial arrangements of workspaces proved vital to
participants’ abilities to effectively manage the information load associated with their
particular types of work. He found that this spatial structuring served a reminding
function for his participants so that they did not have to retain conscious awareness of
discrete pieces of information but instead would structure their environments so that
they encountered the information naturally as needed during their work routines. This
reminding function, he notes, contrasts with the finding fimction in which documents or
other artifacts are carefully classified and organized into discrete information structures,
such as in file cabinets, so that they can found during deliberate searches. Malone
(1983) concludes by noting that “in general, the notion of accessing information on the
basis of its spatial location, instead of its logical classification, is an important feature of
the way people organize their desktops that might profitably be incorporated into

computer-based information systems” (p. 108).

15

Malone’s study initiated a conversation that has continued to inform the design
of user interfaces and to inspire researchers seeking new graphical interfaces for doing
memory work to this day. Barreau and Nardi (1995), for instance, refine Malone’s
conclusions by identifying three basic categories of information on users’ computer
desktops based on temporal and situational needs: ephemeral, working, or archived.
They note that ephemeral and working information is used far more often than archived
information. Consequently, they note that too much attention in interface development
is paid to accessing infrequently-needed archived information at the expense of the
ephemera and working files that users really need in their day-to-day activities. They
conclude by noting that, for ephemera and working files at least, users prefer “a
‘physical’ system in which a specific location is associated with the file, making it more
useful than a purely logical system” (Barreau & Nardi, 1995, p. 41). Thus, they
hypothesize that “users prefer location-based filing because it more actively engages the

mind and body and imparts a greater sense of control” (Barreau & Nardi, 1995, p. 40).

F ertig, Freeman, and Gelemter (1996) counter by arguing that Barreau and
Nardi confuse cause and effect in their assertion that ephemera are more important to
daily work than older information. Fertig, Freeman, and Gelemter (1996) point out that
users would be more likely to regularly use older archived information if their interfaces
gave them readier access to it. Further, they suggest that new, non-spatial metaphors
are needed to help users find and manage larger and larger quantities of stored
information. So, they propose a time-based interface based on the metaphor of a stream

as a better method of managing information over the lifecycle of a project. However,

16

their solution reinforces rather than refutes Barreau and Nardi’s findings regarding

users’ desire for physical rather than purely logical methods for managing memories.

More recently, interface researchers have continued to investigate possible
improvements or alternatives to the search and browse paradigm for both individual and
cooperative work. For example, pervasive computing research like Thayer and
Steenkiste’s (2003) proposes interfaces that integrate physical and digital spaces via
speech and gesture recognition and eye tracking in order to “automate [. . .] much of the
drudgery associated with computers” (p. 82). Zhang and Marchionini (2005) propose a
visualization model that couples the browse and search functions, while Krishnan and
Jones (2005) propose a model of “temporal visualizations” similar to that of Fertig et al.
(1996) to help users manage their information space over time. Krishnan and Jones
(2005) take particular care to emphasize that visualization models for memory work
must be highly customizable by users. They argue that “personal information spaces
are individual. Current systems [. . .] make limited use of features that would
personalize the view so that users’ information space representation is uniquely their
own and lends itself to interpretation,” a finding which again parallels insights from

both the rhetorical tradition and Malone’s research (Krishnan & Jones, 2005, p. 52).

For their part, researchers in computer supported cooperative work have
attempted to demonstrate how creating shared visual and “‘virtual’ physical co-
presence” via digital interfaces can facilitate the sharing of knowledge in teams
(F ussell, Kraut, & Siegel, 2000) as well as to show the inherent difficulties in
coordinating distributed work without the affordances offered by shared information

spaces (Spinelli, Perry, & O’Hara, 2005). This research also demonstrated how “hazy

17

human memory” (Ackerman & Halverson, 1998, p. 46) or “powerful social
mechanisms” (Perry, F ruchter, & Rosenberg, 1999, p. 131) often must suffice when
archives or other memory interfaces break down (Czerwinksi & Horvitz, 2002).
However, as with research on individual desktop interfaces, no consensus has yet been
reached on how to move beyond the limitations imposed on memory work by

contemporary computer interfaces.

Over the past quarter century, computer technology has become the primary
repository of our individual and collective memories, and, as the ecology of computer-
mediated memory technologies grows more sophisticated and powerful — more
graphical, scalable, portable, and embeddable — it increasingly offers the potential to
mediate human memory in ways that transcend the wildest dreams of the forrnulators of
the techniques of the ars memoria. Yet, until our preoccupation with storage capacity
becomes tempered by a similar level of attention to retrieval and use, the ars memoria

will continue to have things to teach us about doing memory.

The Need for a Better Understanding of Memory Work as Rhetorical Practice

These commentaries on the current state of digital interfaces for memory work
reveal the pervasive influence of a particular “memory regime” — a regime that
privileges storage over retrieval (Bowker, 2005, p. 9). According to Bowker (2005),
memory regimes “articulate technologies and practices into relatively historically
constant sets of memory practices that permit both the creation of a continuous, useful
past and the transmission sub rosa of information, stories, and practices from our wild,

discontinuous, ever-changing past” (p. 9). Bowker (2005) further notes:

18

If we completely know a system in the present, and we know its rules of change
(how a given input leads to a given output), then we don’t need to bring to mind
anything about the past [. . .] it remains true that there are modes of remembering
that have very little to do with consciousness. These modes tend to abstract
away individuality [. . .] by substituting rules and constraints on the behavior of

types of people for active recall. (p. 8)

This aspect of a memory regime is mirrored closely in the functioning of a computer
interface, which, when working normally, is intended by its designers to be invisible
and to seem entirely natural to the user. For those interfaces designed for memory
work, this means that the methods of “conventional information retrieval,” like
searching and browsing, become de facto the only methods for doing memory work

with computers (Fast & Sedig, 2005).

While research from archival science and office automation have made
important strides in suggesting new or augmented methods for doing memory work, the
failure to arrive at a viable alternative suggests that new perspectives coming from
different critical traditions is needed as a corrective to the memory regime prevailing in
current symbolic analytic workplaces that relegates retrieval to secondary status in
comparison with storage. For computer interfaces are also, as Selfe and Selfe (1994)
put it, “cultural maps” conveying “the values of our culture—ideological, political,
economic, and educational [. . .] such maps are never ideologically innocent or inert” (p.
485). In its function and effects, the concept of a memory regime as Bowker articulates
it, therefore, equates in many ways to the function and effects of culture. Or, similar to

what Slack and Wise (2005) note about culture, a memory regime performs “the work

19

of selection: the selecting, challenging, arranging, and living of [the] received artifacts
of everyday life” (p. 4). That is, like culture, the memory regime determines what can
and cannot be said, what counts as knowledge, and what evidence can be used as
warrants in arguments. And, also like culture, a memory regime is most readily
discemable at the level of practice, which “denotes a set of socially defined ways of
doing things in a specific domain: a set of common approaches and shared standards
that create a basis for action, communication, problem solving, performance, and
accountability” (Wegner, McDermott, and Snyder, 2002, p. 39). To analyze and
critique a memory regime, then, is in some sense to attempt to bring about cultural
change, and to bring about cultural change, it is first necessary to understand the

practices that arise from this culture with an eye towards changing those practices.

A necessary first step toward a memory regime change, then, entails that we
identify the memory practices of actual writers as they work —— the articulations by
which these symbolic analysts overcome information overload via cunning (and
fleeting) assemblages of both high- and low-technology tools offered by their work
contexts. This dissertation is an attempt to closely observe and understand the complex
scenes of memory work in which writers in contemporary workplaces attempt to
retrieve and adapt stored information to meet the exigencies of rhetorical situations. In
other words, it conceives of memory work as rhetorical practice and investigates the
role in composing and writing of “high” technologies like computer interfaces and
databases in tandem with “low” technological mediations offered by human embodied
interactions in space and time for the purposes of changing memory regimes and

informing the design of more effective technologies in the future.

20

CHAPTER 2: THEORETICAL FOUNDATIONS

Dimensions of a Rhetorical Practice of Memory

The goals of this chapter are to articulate a working description of a rhetorical
memory practice and to ground this description in rhetorical theory and contemporary
psychology in order to arrive at an approach to studying such practices as components
of rhetorical invention. As I noted in the previous chapter, Bowker (2005) coins the
term “memory regime” as a way of describing the collective (or the complete collection
of) memory practices of a given culture. As cultures have subcultures, which both
partake of the features of the larger culture and adapt those features for local
circumstances, memory regimes also have specific instantiations, such as in
organizations, that both adopt and adapt the memory tools and practices of the larger
culture to meet specific organizational needs. Finally, individual memory practices are
the activities and tools by which members of a given memory regime attempt to deal
with information from the past, including a large variety of activities ranging from note-
taking to data-basing. These activities and tools are substantially influenced by the
memory regimes in which they reside but they are also often employed in idiosyncratic
ways in actual work processes based on the experience and cunning intelligence of
individuals as they respond to situations. For example, as the present research study
reveals, some participants handwrite notes in notebooks during meetings so that they
can later refer back to the notes while composing while others do so purely as a method
of imprinting the material more firmly into their own long-term memories -— findings
similar to those of Ann Blair (2004) in her study of note-taking practices throughout

history.

21

However, as Bowker (2005) makes clear, the most important thing about
memory practices is that they exist primarily to enable action in the present rather than
to preserve a perfect record of the past: “one of our chief ways of dealing with the world
is to remember things [. . .] it is one of our chief ways of being in the world as effective
creatures: it is a way of framing the present; a mode of acting” (p. 25). Consequently,
Bowker notes, memory practices are about forgetting as much as they are about
remembering. Building on Derrida’s (1998) theoretical work on archives, Bowker adds
that memory practices tend to be both sequential and jussive. That is, memory practices
are sequential because they partake of standardizing and classifying information so that
it can be found when needed, and they are jussive because they often participate in the
process by which information judged to be of no use is purged from memory — all in

order to enable practical action in the present and future.

Further, according to Bowker, memory practices become incorporated into the
built environments that surround us — our buildings, our workspaces, and our digital
spaces — in other words, in our infrastructures and tools. Alluding to his earlier work
with Susan Leigh Star (Bowker & Star, 1999), Bowker (2005) notes that standards and
classifications, embodied in infrastructures, contain affordances by which we can
offload part of the burden of memory: “we classify in order to be able to forget” (p. 21).
Collectively, these affordances, when they are working properly, enable something
rather like a “standing wave” of memory whereby we are able to manage the
information load necessary to accomplish our daily tasks without becoming
overwhelmed by the work of recall — that is, without becoming overloaded with

information. We are able to do this because infrastructures offer affordances that allow

22

us to offload memory so that precisely the right amount of information required to

perform a given task presents itself at any one moment:

We are not in general able to remember complete stories about the past [. . .]
What we do well is to disaggregate a fact about the past into a number of
standard elements, and then set in train a procedure for reassembling the specific
out of the general. This sets in motion a system of memory recall that is able at

any given moment to create a working version of the past. (Bowker, 2005, p. 18)

In other words, for Bowker, successful memory work depends a great deal on the ability
to respond to kairos, and situational affordances play a large role in this ability:
organizations, Bowker (2005) notes, “delegate memory tasks to the environment” (p.
15). Further, these infrastructures aid memory by organizing and in many cases
limiting what we can and cannot perceive or interact with in a given situation,
phenomena related to cognitive cueing and constraining as noted by psychologists

studying user-centered design (Norman, 2002).

However, Bowker’s (2005) methodology for studying memory is historical and
textual rather than ethnographic or observational, so he focuses more on the larger
systems of memory than on the actual practices by which individuals or groups “set in
train” these procedures for recall during activities (p. 18). While it might, then, be said
that Bowker’s research gives a sense of the where of memory practices, Star’s fieldwork
methods enable her to get a detailed look at the when of memory practices in the real-
time activities of her research participants. From her observations, Star (1999)

formulates the concept of “articulation work” as a label for the “real-time adjustments”

23

that people continually perform below the level of their visible work tasks in order to
make those work tasks and processes flow smoothly (p. 385). Star (1999) elaborates by
describing such articulation work as the invisible “process of assemblage, the delicate
complex weaving together of desktop resources, organizational routines, running
memory of complicated task queues” that goes on below the surface of visible

production work, enabling and supporting it (p. 387). She further notes:

This system is necessarily fragile (as it is in real time), depending on local and
situated contingencies, and requires a great deal of street smarts to pull off.
Small disruptions in the articulating processes may ramify throughout the
workflow of the user, causing the seemingly small anomaly or extra gesture to
have a far greater impact than a rational user-meets-terminal model would

suggest. (Star, 1999, p. 387)

Star does not describe them as such, but many of these fragile and impromptu practices
of articulation work are, in fact, memory practices, the fragile and fleeting assemblages
by which we continuously attempt to preserve our standing wave of memory through
space and time. Further, many of these memory practices do rhetorical work because
they support and partake in the activities of invention in which we utilize contextual
affordances to retrieve and manipulate stored information in timely and appropriate

(i.e., kairotic) ways.

The Role of the Body in Writing and Memory

Bowker (2005) and Star (1999) provide a good framework for understanding

memory practices as occurring at the confluence of spatial and temporal contexts — the

24

where and when of infrastructures, in other words). What they are less interested in
doing is exploring the role of bodies in these articulating processes, which perhaps
might be thought of as the who, what, and how of infrastructures. However, as I will
show, the rhetorical tradition, at least as it was originally formulated under the Greeks
and Romans, understood memory as fundamentally psycho-somatic, involving the
actions of both the mind and the body. Therefore, if we want to better understand
memory practices, particularly rhetorical memory practices, it is important that we pay
attention to the embodied aspects of human interactions with infrastructures in space
and time. That is, the rhetorical tradition points to the fact that only by accounting for
the activities of memory performed by body and mind in tandem can we fully
understand these activities as culturally determined practices, as manifestations of
particular memory regimes. An obvious place to look for manifestations of these

memory “street smarts” is with writers (Star, 1999, p. 387).

The purpose of this study, then, is to explore the ways in which writers rely on
their embodied senses in order to perceive and use the infrastructural affordances of
their workplaces to help them avoid information overload while inventing and
composing. This is an innovative approach to understanding memory because studies
of the role of the senses in mediating memory in complex “open-ended design
process[es]” like writing have in the past tended to become preoccupied with only one

of our embodied senses: our sense of sight (Sharples, 1996, p. 127).

This preoccupation with the role of sight in mediating composing tasks goes
back at least to Cicero (1942), who calls sight “the keenest of all our senses,” and so it

is not really surprising that many writing process studies focus on visual aspects of

25

composing to the exclusion of other ways in which we perceive the world (p. 469).
Sharples (1996), for instance, points out that “one way to overcome the difficulties of
performing such complex knowledge manipulation in the head is to capture ideas on
paper [. . .] in the form of external representations that stand for mental structures. So
long as ideas, plans, and drafts are locked inside a writer’s head, then modifying and
developing them will overload the writer’s short-term memory” (p. 135). In making this
assertion, Sharples is, of course, reinforcing the findings of a long line of writing
process researchers like Flower and Hayes (1981) who note that the ability to see one’s
own words on paper helps a writer compose: “the logic which moves composing
forward grows out of the goals [that] can be both sustained and influenced by [. . .] the
text itself” (p. 380). Similarly, Neuwirth and Kaufer (1989) note that, when composing,

“external representations then can be useful for keeping track of goals” (p. 328).

Perhaps no one has endowed the sense of sight with so much importance in
thinking and composing as Walter Ong. According to Ong (1982), writing
“transformed human consciousness” (p. 78) because it moved articulated thoughts (i.e.,
speech) out of the “evanescent” (p. 32) realm of aural sensation and into “a new sensory
world, that of vision” (p. 85). This new visual practice of communication enabled, Ong
says, more abstract and complex forms of thought. Ong (1982) puts it this way, “by
taking conservative functions on itself, the text frees the mind of conservative tasks, that
is, of its memory work, and thus enables the mind to turn itself to new speculation” (p.
41). Sight can do this because “sight isolates” while “sound incorporates” which gives
the thinker/composer perspective on his or her thoughts, enabling, it seems, more

abstract and conceptual thought (Ong, 1982, p. 72).

26

Cicero and Ong are, of course, correct: sight is undeniably important to
composing and is probably, as they claim, the sense we rely on most when writing. Yet
still, as cognitive research shows, we miss something important if we become too
preoccupied with the sense of sight to the exclusion of the other ways in which our
bodies reason and remember. For example, as the following interview transcript taken
from Linda Flowers’ “Cognition, Context, and Theory Building” (1989) demonstrates,
research into the supposedly ‘situated’ contexts of student writers tends to occlude the

fully embodied nature of their composing practices:

Ron: I try to write [an assignment] as soon as I can and let them look at it.
Even take it right to the teacher, and say, look at this. Am I going in the right

direction or not?

Interviewer: That’s a kind of expensive way to do it, isn’t it?

Ron: You pick up things. You pick up good things. It’s expensive in terms of
that paper, but it’s not expensive in terms of putting that away for future
reference. [. . .] It’s not really a conscious process that I go through. You just
got to listen. I don’t know if it sounds weird or what. But I sit there and I watch
them during the lecture, I listen to key words that they use. They register. (p.

293)

Flower (1989) uses this section of transcript to highlight the “savvy [. . .] and highly
intentional effort” that students such as her interviewee exert in order to understand and
interpret a given writing context (p. 293). What she ignores in her discussion, however,

is the physical and metaphoric language this savvy student writer uses to describe his

27

learning and writing processes. The student himself probably does not consciously
realize it, but his words reveal that writing and the memory-work associated with
writing are physical as much as mental events, and that this physicality is derived not
just from what he sees (“I watch”) but what he hears (“I listen”) and feels (“I sit,” “you
pick up,” “going in the right direction”). Even when physically confined by a
classroom desk, the student seems to be thinking with his body, not just with his mind:
his sensations while sitting at his desk influence what he absorbs into memory, how
those memories are deployed in composing, and, of course, how his composing
practices are recalled and articulated in the interview situation itself. In his
retrospective account, the student seems to be using the space around him to help him

remember what he needs to remember to accomplish his task.

Lakofl and Johnson shed some light on why Flower’s student seems to be
thinking with his body. They point out that “there is no such fully autonomous faculty
of reason separate from and independent of bodily capacities such as perception and
movement [. . .] reason uses and grows out of such bodily capacities” (Lakoff &
Johnson, 1999, p. 17). Metaphors like the ones the student uses to describe his
composing processes arise out of his bodily experience. In fact, according to Lakoff
and Johnson (1999), virtually all metaphors are based on bodily experience: they exist
so that the “inferential structures of concrete domains” can be “employed in abstract
domains.” (p. 155). So, the student composes with his body and subsequently
articulates this composing process in bodily language because there is simply no other
way he could do it: “can we think about subjective experience and judgment without

metaphor? Hardly” (Lakoff & Johnson, 1999, p. 59). Further, according to Lakoff and

28

Johnson (1999), such use of metaphor is “the principal instrument of abstract reason,”
again, because they are the means by which concrete embodied knowledge can be
employed in abstract reasoning (p. 155). Lakoff & Johnson’s chain of logic, then, runs
something like this: we think via metaphor, most metaphors (“primary metaphors”) are

derived from “sensiromotor domains,” therefore we think with our bodies (1999, p. 45).

The claim for the centrality of metaphor in thinking is, of course, not at all new
in rhetorical theory. Catherine Hobbs (2002), for example, points out that in the work
of Aristotle and Vico, “metaphor is a cognitive instrument more than an ornament,
productive of new knowledge for the individual and the culture” (p. 71). So, if we
accept Lakoff and Johnson’s assertion that metaphors almost always arise from the
body, there should, in theory, be a very strong tradition in rhetoric and writing studies
related to how we use the body to compose, yet this has not been the case. Kristie
Fleckenstein (1999) notes this absence: “bodies as sites of and participants in meaning-
making have been elided [. . .] we need an embodied discOurse [that] locates an
individual within concrete spatio-temporal contexts” (p. 281). Like Lakoff and
Johnson, Fleckenstein (1999) also links the body to metaphor and memory in
composing: “operating according to metaphoric [. . .] logic, corporeal texts are the
means by which we carry our bodies in our minds [. . .] corporeal codes stabilize
discursive codes and produce a language from pulse beats, memories, and images” (p.
290). Finally, Fleckenstein (1999) urges the field of rhetoric and writing to
acknowledge the corporeality of texts as one method of “refiguring writing, teaching,

and researching in composition studies” (p. 298).

29

In fact, several influential researchers have attempted to answer F leckenstein’s
concerns over the past few years. Beverley Sauer (1998, 2003), for instance, explores
miners’ use of physical gestures derived from their procedural memories of embodied
experience in the meaning-making process: “gesture is both an iconic image and an act
of rhetorical meaning-making that assists and constructs an individual’s knowledge of
risk” (2003, p. 257). Similarly, Haas and Witte (2001) examine the role of embodied
actions like movements and gestures as “pre-texts” that mediate the thought processes
of blue-collar city workers during their interactions with white-collar engineers while
revising a report (p. 444). Likewise, although he is not necessarily focusing on
composing in the narrow sense of putting pen to paper, Mike Rose (2004) studies the
role of embodied actions in shaping the thought processes that lie behind such mundane
yet sophisticated activities as waiting tables or carpentry. Finally, Prior and Shipka
(2003) speculate about how the buzzing timer on a clothes dryer and the subsequent
activity of folding laundry affect the composing processes of a college professor as she
writes a scholarly article while working at home. With the exception of Prior and
Shipka, however, all of these studies operate on what could be labeled a “deficit model”
of the role of the body in composing. That is, none of them really focuses on the
embodied aspects of composing as hallmarks of expertise in writing, as I believe the
rhetorical tradition calls for. Instead, they seem to conceptualize the employment of
embodied senses, particularly senses not derived from sight, as crutches or as the
remediation strategies used by less-skilled writers attempting to express their meanings

in language.

30

It seems, then, that we have lost something somewhere along the way in the
history of rhetoric: we have lost the idea that the expert writer is the writer who
practices a physical as well as a mental art. What is needed then is a renewed theory of
embodied rhetoric grounded in contemporary cognitive theoretical constructs that lend
themselves to empirical study. Such an embodied rhetoric of memory will, by
highlighting the expertise of writers in organization as the product of both a physical
and mental discipline and by revealing the working environment in which this expertise
is developed, contribute valuable insights that may influence how technical
communication is studied and taught and the physical, digital, and social spaces in

which it is practiced.

Aristotle’s Embodied Epistemology

While Aristotle does not offer the sort of tidy summation of the link among the
body, writing, and memory that Plato provides at the end of Phaedrus, his thought is
indispensable for understanding memory and writing as embodied activities. Although
his theories about the body, thinking, language, and memory are dispersed throughout
his works, Aristotle principally articulates them in his treatise “On Memory and
Reminiscence,” which Murphy (2002) points out provides part of the “general theory of
human action” that undergirds Aristotle’s theory of rhetoric (p. 213). First, Aristotle
and his medieval successors like Aquinas believed that the body and its sense
perceptions played a far greater role in thinking than has post-Cartesian thought. Theirs
was a fundamentally embodied epistemology. That is, Aristotelians like Aquinas held
that “the activity of thinking and the activity of having a sense perception are

fundamentally analogous, not fundamentally different” (Carruthers, 1990, p. 57). Or, as

31

Lakoff & Johnson (1999) articulate it, Aristotle “locates reality ultimately in the world,
and he thus sees our thought as dependent upon the nature of the world. [. . .] Thus, for
both Plato and Aristotle, there is no separation between the mind and the world” (p.

374).

As for the role of memory in thinking, it too involves the whole body. In one of
the most important but confusing parts of “On Memory and Reminiscence,” Aristotle

states:

One might be puzzled how, when the affection is present but the thing is absent,
what is not present is ever remembered. For it is clear that one must think of the
affection, which is produced by means of perception in the soul and in that part
of the body which contains the soul, as being like a sort of picture, the having of
which we say is memory. For the change that occurs marks in a sort of imprint,
as it were, of the sense-image, as people do who seal things with Signet rings.

(Sorabji, 1972, p. 50)

At first glance, Aristotle seems to be saying that we only remember in pictures, which,
if true would, according to Virginia Allen (1993), be a faulty notion because, “hasty
introspection reveals that our knowledge of such things as typing, playing the guitar,

and driving a car are not mediated with images” (p. 51).

Yet, Allen is missing a key nuance here. Aristotle is not saying that these are
exclusively or literally visual images but, rather they are “like a sort of picture”
(Sorabji, 1972, p. 50, emphasis mine). Carruthers and Sorabji avoid Allen’s mistake by

labeling these “quasi-imprints” phantasms to distinguish them from literal visual images

32

(Carruthers, 1990, p.16; Sorabji, 1972, p. 14). Carruthers ( 1990) glosses the phantasm
this way: the “phantasm is the final product of the entire process of sense perception,
whether its origin be visual or auditory, tactile or olfactory. Every sort of sense
perception ends up in the form of a phantasm in memory” (p. 17). In other words,
rather than exclusively (or even mostly) a visual picture, a memory is more like a
multirnodal “snapshot” derived from all our embodied senses in a given moment in
space and time: “all mnemonic advice stresses the benefits to be gained from forming
memories as ‘scenes’ that include personal associations [. . .] the need to impress the
circumstances during which something was memorized [. . .] how one feels, the gestures
and appearances of one’s teacher, the appearance of the manuscript page, and so on”
(Carruthers, 1990, p. 60). This interpretation of the Aristotelian phantasm as a “scene”
perceived from the embodied perspective of the rememberer is reinforced by Sorabji
(1972), who says “Aristotle seems to imply [that] that the memory-image is a copy of
one’s view of that scene” (p. 7). Murphy (2002) concurs with Sorabji, adding that data

from the other senses is “collated” in the phantasm (p. 218).

The process of deliberately recollecting memories (as opposed to simply random
recalling), then, entails finding or locating these phantasms/scenes/snapshots either via
repeating some aspect of the physical circumstances in which the original sense
impression occurred (e. g., walking to the foyer and retracing your steps to try to figure
out what you did with your keys when you walked into your house) or by using some
sort of artificial heuristic technique like the place/image mnemonic. The recollection
process also involves an act of reconstructing and interpreting the embodied sensations

laid down during the original experience of the thing being recalled: “recollection was

33

understood to be a re-enactment of experience which involves cogitation and judgment,
imagination and emotion” (Carruthers, 1990, p. 60). Since, in most cases we can’t
actually recreate the exact physical circumstances of the original experience (the
memory work involved in composing tasks is rarely as simple as finding one’s keys),
artificial methods (that is, methods that are susceptible to training or the product of an

art) and tools for recollection become vital.

Yet, recollection during open-ended design tasks like composing can be difficult
and can resemble the cognitively demanding, labor-intensive process of dredging up
declarative memories one-by-one that experts know how to avoid. To make the process
of recollection easier, Aristotle offers a number of possible methods for tapping into and
manipulating phantasms for purposes of deliberately recollecting memories during
composing tasks: employing the topoi as “organizing modes of recollection” helps us to
envision ourselves in relation to the points we are trying to make by “initiat[ing]
memory in certain directions” (Murphy, 2002, p. 220); using metaphor, as I’ve already
mentioned — and as the example dialog from Flower’s study demonstrates — helps us
connect “concrete domains” to abstract ones (Lakoff & Johnson, 1999, p. 155);
“tagging memory emotionally” makes remembering easier by adding an internal
sensations to the external physical sensations of the phantasm snapshot (Carruthers,
1990, p. 60); and, perhaps most importantly, repeating and practicing a particular
composing activity habituates us to the memory demands of particular rhetorical
situations (Murphy, 2001 & 2002, points out that the Romans made habit the

foundation of their educational system).

34

The Aristotelian Phantasm and Social Theories of Cognition

Each of the methods for engaging memory during composing that Aristotle
offers makes intuitive sense as means of performing memory work, but how does the
phantasm fare as a construct in light of contemporary developments in cognitive
psychology, and, more pertinently, how might we study similar methods in the actual
practices of contemporary writers as they work? Answers to these questions can be
found in recent theories of social cognition deriving from activity theory as formulated
by the great Soviet psychologists Lev Vygotsky and Alexei Leontiev. In fact, activity
theory, with its focus on external physical actions as keys to understanding internal
mental states and its assertion that “perception is an integral part of human interaction
with the world” have gone a long way towards refuting the legacy of post-Cartesian
understandings of human thought as a product of an a-social, a-historical, a-material,
and disembodied mind and, consequently, towards revalidating Aristotle and his

successors’ embodied epistemology (Kaptelinin & Nardi, 2006, p. 81).

Of particular importance in this endeavor are theories of situated cognition
which posit that the activities and contexts in which learning occurs are inseparable
from and co-productive of knowledge itself. First, consider the following passage in
which Brown, Collins, and Duguid (1989) articulate the central construct “knowledge”

lying behind their theories about how people learn in and through situated activities:

Knowledge, we suggest, similarly indexes the situation in which it arises and is
used. The embedding circumstances efficiently provide essential parts of its
structure and meaning. So knowledge, which comes coded by and connected to

the activity and environment in which it is developed, is spread across its

35

component parts, some of which are in the mind and some in the world much as

the final picture on a jigsaw is spread across its component pieces. (p. 36)

It is not too much of an overstatement to suggest that the construct that Brown, Collins,
and Duguid are describing here serves essentially the same purpose in remembering and
drinking as the Aristotelian phantasm. That is, their “final picture,” like the phantasm,
is something similar to an embodied snapshot linking interior phenomena —- knowledge
— inextricably to the external circumstances of place and time in which that knowledge

was learned — in which it entered memory (Brown, Collins, and Duguid, 1989, p. 36).

Subsequent theorizers of situated cognition add nuance to this construct. For
instance, Reynolds, Sinatra, and Jetton (1996) describe situated cognition in this way:
“situated cognition [. . .] attempts to account for how one learns in a conceptual
environment. The conceptual environment consists of the external world as perceived,
the internal representations of the perceptions, and the resulting interactions” (p. 100).
These “internal representations” seem quite similar in both origination and in function
to the phantasm. Cybernetic theorists like Clancey (1997), who have turned to situated
cognition as a means of formulating new approaches to artificial intelligence, add
further weight to this claim. Clancey (1997) says “conceptual knowledge, as a capacity
to coordinate and sequence behavior, is inherently formed as part of and through
physical performances. The formation of perceptual categorizations and their coupling
to concepts provides material for reasoning (inference), which then changes where we
look and what we are able to find” (p. 5). This “coupling” of perceptual information

with declarative conceptual knowledge that Clancey recognizes, again, seems to

36

ftmction in much the same way that the phantasm does by collating diverse sensory

inputs into usable and recollect-able scenes which drive human thinking.

Moreover, some of the methods of tapping into and manipulating these “final
pictures” or “internal representations” that the situated cognitivists posit as critical to
achieving expertise in a given task domain resemble the advice offered by Aristotle for
tapping into the phantasm. Habit, in particular, plays a critical role for both Aristotle
and the situated cognitivists. According to Aristotle (1952), “acts of recollection, as
they occur in experience, are due to the fact that one movement has by nature another
that succeeds it in regular order. If this order be necessary, whenever a subject
experiences the former of two movements thus connected, it will [invariably],
experience the latter” (p. 693). Murphy (2002) points out that what Aristotle is
describing is habit, noting that “the tendency to act in a certain manner, derives from
memory in that unrecollected choices create a potential motion of the soul in advance of

recollection” (p. 218).

To deliberately trigger these “unrecollected choices” during some deliberate task
like composing, then, one needs to practice so that the action becomes habitual during
subsequent performances of the task: “accordingly, therefore, when one wishes to
recollect, this is what he will do: he will try to obtain a beginning of movement whose
sequel shall be the movement which he desire to reawaken” (Aristotle, 1952, p. 693).
Or, as Carruthers (1990) glosses Aristotle’s notion of habit articulated in the
Nichomachean Ethics: “one’s hexis or habitus is developed by the repetition of
particular emotional responses or acts performed in the past and remembered, which

then predispose it to the same response in the future. [. . .] Experience is made from

37

many repeated memories, which in turn are permanent vestiges of sense perceptions”
(p. 68). Carruthers (1990) links the role of the phantasm to the laying down of habit: “it
is the spatial, somatic nature of memory-images that allows for secure recollective
associations to be formed [. . .] because it is also a physiological process, recollection is
subject to training and habituation in the manner of all physical activity (p. 63). Finally,
Sorabji (1972) corroborates Carruthers by stating, “it looks as if Aristotle’s view is that,
whenever images regularly follow each other, this is by way of habit. The habit may
have become established either because the images were naturally fitted to occur in a
certain order, or [. . .] as a result of artifice [i.e., training]” (p. 45). Habit, of whatever

type, is for Aristotle a product of physical as well as mental and emotional training.

Quite similarly, Brown, Collins, and Duguid (1989) note that “understanding is
developed through continued, situated use” (p. 33). From this idea, they formulate the
notion of the “cognitive apprenticeship” as the ideal method of learning to be an expert
in a given task domain. Again, like Aristotle, the process of habituation achieved

through extended apprenticeship is as much physical and emotional as it is mental:

Cognitive emphasizes that apprenticeship techniques actually reach well beyond
the physical skills usually associated with apprenticeship to the kinds of
cognitive skills more normally associated with conventional schooling. This
extension is not as incompatible with traditional apprenticeship as it may at first
seem. The physical skills usually associated with apprenticeship embody
important cognitive skills, if our argument for the inseparability of knowing and

doing is correct. (Brown, Collins, & Duguid, 1989, p. 39)

38

So, a cognitive apprenticeship is a means of achieving a hexis. The fundamental
congruence of the Aristotelian conception of knowledge and that of the situated

cognitivists leads them to articulate quite similar theories of learning.

Two other methods for manipulating the phantasm that Aristotle offers — the
topoi and metaphor — suggest yet another activity theory-inspired cognitive theory:
distributed or joint cognition. Building on insights from Edwin Hutchins’ seminal study
of ship navigators in Cognition in the Wild (1995a), EngestrOm and Middleton (1996)
identify the central construct of knowledge lying behind theories of joint cognition in

this way:

[The] unit of analysis [is] a culturally constituted functional group rather than an
individual mind. This theory reconceptualizes ‘inforrnation’ as the propagation
of representational states of mediating structures that make up the dynamic and
substance of any complex system. These structures include internal as well as

external knowledge representations, (knowledge, skills, tools, etc.). (p. 6)

These information “structures” fill the same role in theories of joint cognition as
phantasms do in Aristotle’s theory of memory. Consequently, as this passage suggests,
phantasms, like these information structures, exist not only inside the head of the
individual but also in the collective, in the “functional group.” That is, because of the
shared spaces in which we live and work and the commonality of our embodied
experiences in these spaces, the Aristotelian phantasm, as a snapshot of sensory
experience, is inescapably both an internal and an external representation of knowledge.

Viewed in light of joint cognition, Aristotle’s phantasm is a theory of joint as well as

39

individual memory because it offers an explanatory framework for collective activity, as
the internal representations which constitute the phantasm are propagated across
individuals through language, through spaces, through tools, and through artifacts.
Metaphors and topics, then, are methods not only for individual understanding and
drinking but also for joint thinking because they are methods for sharing and

communicating knowledge representations.

To take another example: Describing their case study of joint cognition in
airport workers, Goodwin and Goodwin (1996) point out that “in these data we are able
to catch a glimpse of the social and historical processes through which a community
accumulates experience of the habitual scenes that constitute their working
environment, and articulates for each other how these scenes should be properly
interpreted” (p 83). These “habitual scenes,” then, give rise to the metaphors and
determine the common topics or “places” which a community uses to recollect and to
reason. As I note above, Murphy (2002) points out that the topoi “initiate memory in
certain directions” (p. 220) Metaphors and topics, then, depend on the “embeddness of
knowledge” that is, they assume that “the ability to see something is always tied to a
particular position encompassing a range of phenomena including placement within a
larger organization, a local task, and access to relevant material and cognitive tools”
(Goodwin & Goodwin, 1996, p. 61). The airport workers are able to work together
successfully because of the shared contexts in which their memories were originally laid
down. By virtue of their training and common experience, they share phantasms, or, as
Fentress and Wickham (1992) put it, social memory is strongest “at the level of shared

meanings and remembered images” (p. 59).

40

Finally, one particularly important aspect of activity theory — its focus on tool
use — helps us understand how Aristotle’s embodied epistemology may have been
employed at the level of practice by the ancient rhetoricians. Activity theory posits that
tools operate at both external (embodied) and internal (psychological) levels, bridging
the body and the mind in everyday practice: “tools shape the way human beings interact
with reality [and] the shaping of external activities eventually results in the shaping of
internal ones” (Kaptelinin & Nardi, 2006, p. 70). Tools also function as enablers of joint
cognition by bridging the knowledges of individuals: “tools usually reflect the
experience of other people who tried to solve similar problems earlier and invented or
modified the tool to make it more efficient and effective [. . .] the use of tools is an

accumulation and transmission of social knowledge” (Kaptelinin & Nardi, 2006, p. 70).

Memory tools are, of course, no different from any other tool: they are socially
acceptable aides to natural memory that often become internalized and are often
employed in idiosyncratic ways by individuals during the course of actual practice. The
function of a memory tool might then be thought of as an aide to maintaining the
standing wave of memory that enables effective action in the present by providing just
the right information about the past. The memory tool accomplishes this because it
contributes to the sequential and jussive work of the memory practice. The walking
mnemonic of the Roman rhetoricians was, of course, just such a tool — it enabled
Roman rhetoricians to think on their feet in the Roman forum by helping preserve the
order of their speeches and by ensuring that only the correct information would be “in
sight” and therefore in mind at the correct time that it was needed. That is, by turning

difficult declarative memory tasks like recalling the content of one’s speech into easier

41

procedural or natural memory tasks like walking down the street, the walking
mnemonic might be thought of as functioning as a sort of embodied simulation or a
filmstrip version of Aristotle’s phantasm, in which the orator imagines him or herself
interacting with places derived from habitual experience. As Luria’s (1987) case study
in The Mind of a Mnemonist demonstrates, the walking mnemonic is an incredibly
effective tool for doing memory work, but, most importantly for the present discussion,
it was originally theorized from an embodied epistemology that fully enlisted the body
— its perceptions and sensations — in its memory practices. As a tool for memory, then,
the walking mnemonic will doubtless have a longer shelf-life than many of the digital
memory solutions detailed in the previous chapter’s consideration of archival

technology and office automation.

This study, therefore, will attempt to better understand the memory “street smarts”
of writers by exploring the ways in which writers use and adapt the perceptual
affordances offered by existing tools and infrastructures to do the memory work of

composing (Star, 1999, p. 387). To accomplish this, the study asks these questions:

0 Do writers use the affordances of their embodied contexts to support memory-

related demands of composing?

o Where/when in the composing process do embodied-contexts seem to be most

important?

0 How, if at all, do embodied contexts aid composers’ memory work during

invention?

42

CHAPTER 3: METHODOLOGY AND RESEARCH METHODS

Scenes of Memory Work

Studying memory is a complex undertaking. In one sense, memory is
ubiquitous because memory as stored information impacts virtually everything we do
and the various aides to memory are legion, from our email in-boxes to reminders
posted on our bulletin boards, to our Internet browser bookmarks, to the organization of
papers on our desks. In another sense, of course, memory is ephemeral, often taking
place inside our heads in fleeting moments of deliberate recall or unintentional
recollection. Approaches to studying the function and effects of memory in human
activities, such as writing, are similarly diverse, ranging from broad explorations of
memory as social narrative in Fentress and Wickham (1992) to narrowly-focused and
quasi-experimental attempts to describe the cognitive processes of memory at work
during various activities, such as the composing research of Flower and Hayes (1981)

and Kellogg (1996).

Bowker is critical of both extremes. As opposed to broad studies of narrative
and intertextuality, Bowker (2005) encourages a more situated focus by pointing out
that “stories are told in a context, under a description” (p. 7). Cognitive approaches are

even more misguided, he says:

Memory is often, and wrongly, conceived of as an act of consciousness and
associated with what can be called to mind. By this light, it is often seen as the
act of deciphering traces from the past. We don’t analyze the movement of

icebergs by studying the bit that appears above the surface of the sea; nor should

43

we study memory in terms of that which fires a certain set of neurons at a
determinate time. We as social and technical creatures engage in a vast span of
memory practices, from the entirely non-conscious to the hyperaware. (Bowker,

2005,p.8)

Bowker’s study, with its extrapolation of work practices based on historical analysis,
presents one possible middle approach between the two poles. However, rhetorical
theory offers a refinement and additional level of specificity to such a middle approach
because, as I argue in Chapter 1, a rhetorical theory foregrounds the notion of kairos
over chronos in any study of memory-in-use or memory as practice — a move that

Bowker’s methodology does not allow him to make.

John E. Smith’s (2002) distinction between chronos time and kairos time helps

make clear the importance of rhetorical theory to memory study:

Chronos [. . .] means the uniform time of the cosmic system [. . .] In chronos we
have the fimdamental conception of time as measure, the quantity of duration,
the length of periodicity, the age of an object or artifact, and the rate of
acceleration of bodies. [. . .] The questions relevant to this aspect of time are:
‘how fast?’ ‘how frequent?’ ‘how old?’ and the answers to these questions can
be given in cardinal numbers. By contrast, the term kairos points to a qualitative
character of time, to the special position an event or action occupies in a series.
[. . .] The question especially relevant to kairos is ‘When?’ ‘At what time?’ (p.

47)

44

That is, where both macro- and micro-level approaches to studying memory (as well as
approaches to designing office automation and archival technology) tend to focus on
memory as a matter of storage — of where and how to save information (i.e., the
sequential aspect of the archive in Derrida’s scheme) — rhetorical theory focuses on
context and use. Consequently, the concerns of a chronos-centered view of memory lie
primarily with the accumulation of information, with concerns such as how information

is stored in databases, in file systems, in the structures of the organic brain.

A kairos-centered theory of memory, on the other hand, entails that we focus on
both storage and recall in specific times, places, and, most importantly, in order to
achieve specific purposes — in other words, on the rhetorical context of memory.

Sheard (1993) makes this connection between kairos and context explicit:

Kairos is the ancient term for the sum total of ‘contexts,’ both spatial (e.g.,
formal) and temporal (e.g., epistemic), that influence the translation of thought
into language and meaning in any rhetorical situation. Kairos encompasses the
occasion itself, the historical circumstances that brought it about, the generic
conventions of the form (oral or written) required by that occasion, the manner
of delivery the audience expects at that time and place, their attitudes toward the
speaker (or writer) and the occasion, even their assumptions about the world

around them, and so on. (p. 292)

Sheard’s conceptual linking of kairos and context echoes the work of Kinneavy (1986,
2002). Kinneavy (1986) identifies kairos as a critical construct tying rhetorical theory

to empirical research: “the concept of situational context, which is a modern term for

45

kairos, is in the forefront of research and thought in many areas” (p. 83). In particular,
Kinneavy calls attention to the compatibility between theories of kairos and domains Of
inquiry, such as ethnomethodology, that focus on the ways in which context is “defined
dynamically” in a given situation (Dourish, 2004, p. 22). Finally, Kinneavy (1986)
notes the ongoing relevance of rhetorical theory to empirical research designs: “it may
be that modern treatments of situational context can learn something from the handling

of the same topic in antiquity. I would argue that they can” (p. 85).

So, then, how does one design a study of contextualized memory use that is
informed by rhetorical theory? Rhetorical theory first provides the notion that rhetorical
situations, encompassing definable speaker/writer, audience, and purpose, are the places
and times to look for this contextualized memory use. Further, as Chapter 2
demonstrated, classical rhetorical theories of memory are built on an embodied
epistemology largely validated by contemporary cognition research, so such research
must also account for the role of the body in the thinking and remembering that occurs
in rhetorical situations. Consequently, the central methodological aim of this study of
memory in composing is to observe and account for the embodied contexts of
composing situations and to relate these as closely as possible to the activities of

memory in those situations.

Rhetorical theory itself provides a label for and approach to bounding such a
contextualized inquiry in the notion of a “scene.” Much like a scene in a drama, a scene
of rhetorical memory work might provisionally be defined as the lamination of
otherwise spatially and temporally discrete situations inhabited by particular actors,

both human and non-human, in which the activities of storing information are linked to

46

the activities in which that stored information is retrieved and used (see Figure 1).

These laminations are discemable chiefly through the rhetorical purposes of the actors

Figure l: The scene of memory work

Rhetorical purpose/
situational exigency

/\ boundsthe scene

The scene of
memory work

 

   
   

  
   

Activity of
retrieval/ use

 

Activity of
storage

 

 

 

 

Time =

involved but they are also responsive to the situational exigencies arising from the times
and locations and spaces and places in which the memory activities occur. Thus the
notion of kairos as both context and occasion — rhetorical purpose and situation of use —

links the activities of storage and retrieval in the scene of memory work.

Studying kairotic memory as scenic has ample precedent. First, such an
approach is in keeping with Aristotelian embodied epistemology, as Carruthers (1990)

points out:

47

All mnemonic advice stresses the benefits to be gained from forming memories
as ‘scenes’ that include personal associations [. . .] the need to impress the
circumstances during which something was memorized [. . .] how one feels, the
gestures and appearances of one’s teacher, the appearance of the manuscript

page, and so on. (p. 60)

A version of the classical understanding of scene recurs in modern rhetorical theory in
Kenneth Burke’s pentad, which posits the scenic nature of all human communication.
According to Sheard (1993), “Burke describes a ‘scene’ as the container for acts and
agents” (p. 305). Finally, as with the Aristotelian phantasm, contemporary cognitive
theory points to the embedded-ness of human thought and activities in particular scenes.
According to cognitive scientist David Kirsh (1995), “in having a body, we are spatially
located creatures: we must be facing some direction, have only certain objects in view,
be within reach of certain others. How we manage the space around us, then, is not an
afterthought; it is an integral part of the way we think, plan, and behave” (p. 31). Kirsh
(1995) cites our use of space as particularly important to memory in use: “space is a
resource that must be managed, much like time, memory, and energy. When we use
space well we can often bring the time and memory demands of our tasks down to
workable levels” (p. 32). In short, the props and sets of the scenes in which memory
work occurs play a key role in the kairos of the scenes. As the ancients recognized, the

kairos of a scene arises at least in part from the physical infrastructure of the scene.

Labeling a situation of memory work a “scene,” then, attempts to account for the
spatial and temporal aspects of activity in Bowker’s (2005) memory practices and Star’s

(1999) articulation work: the scene is the “where” and “when” of memory

48

infrastructures determined by their use in fulfilling rhetorical purposes. These time and
place combinations encompass the spaces, tools, people, and situations in which
rhetorical invention occurs. In short, they are where memory regimes, infrastructures,
and bodies come together in memory activities that further the rhetorical purposes of the
actors involved. The scene of memory work, then, represents perhaps the chief
contribution of rhetorical theory to studying memory since it inextricably binds
activities of storing memories to the activities of retrieval in which those stored
memories are used to achieve rhetorical purposes in specific situations. Without such
binding, inquiry into memory practices, as the review of research from archival science
and office automation demonstrates, tends to become preoccupied with the tools and
practices of storage at the expense of the tools and practices of use. Because it makes
explicit the connection between storage and use, the scene, therefore, would seem to be

a better focus of inquiry into memory practices.

As for how to go about studying the memory work in a scene, a couple
possibilities present themselves. Burke’s (1969) method of analyzing rhetorical
situations according to pentadic ratios of act, scene, agent, agency, and purpose—
dramatism—presents a method that lends itself well to the study of texts. More
appropriate to the present study of practice, however, are constructs from the various
branches of social cognition theory, detailed in Chapter 2, that confirm and validate
earlier rhetorical theories of memory in practice. These provide a language for
describing and methods for studying scenes as temporal and spatial laminations in
which memories are recalled through embodied perceptions and enacted in observable

activities and interactions. Activity theory suggests that we pay attention to tools and

49

artifacts; situated cognition theory suggests that we focus on place and space; and joint

cognition theory requires that we look at the role of other actors, human and nonhuman.

To understand the relationship of these categories of visible objects (tools,
artifacts, place, space, and bodies) to the operations of memory, a further category of
data must be collected from a scene, that of the activities that take place in/during the
scene. Ultimately, the central premise of this methodology is that observation of the
visible aspects of the scene is the key to finding memory in the scene. As activity
theory posits “Activity theory maintains that internal activities cannot be understood if
they are analyzed in isolation from external activities, because there are mutual
transformations between the two kinds of activities. Intemalization is the
transformation of external activities into internal activities” (Kaptelinin & Nardi, 2006,

p. 69).

The Research Site

While the notion of a scene provides the methodological and theoretical bounds
for data collection, the selection of a research site was equally important. Where the
notion of the scene provides a framework for studying instances of memory regimes in
the intersection of infrastructures and people in time, the research site provides a
location and an opportunity for studying these intersections in shared space — physical,
digital, and social. In other words, the research site is the stage on which the scenes of
memory work are enacted, a site which is both constitutive of and constituted by the
memory regimes at work in the larger society.

In keeping with the aim of this research project to understand the memory

activities of writing as symbolic-analytic work, therefore, I selected a knowledge-

50

intensive workplace as the research site. This site was a privately-owned medium-sized
(around 175 employees) commercial software firm in the US Midwest. The firm, which
I will call “Software Unlimited” throughout this dissertation, creates and markets image
manipulation and screen recording software to consumer, corporate, and university
customers in the US and through resellers in 30 other countries. As part of its
increasing focus on international expansion, the company’s products, including
software and documentation, are localized into five languages. At the time of the study,
Software Unlimited was over twenty years old and, like other software firms of similar
age, had experienced periods of growth and periods of downsizing; however, during the
two years immediately prior to the study, the company had undergone tremendous
growth, nearly tripling its number of employees.

The research described in this dissertation took place in Software Unlimited’s
headquarters offices. These headquarters offices, where most of the company’s
employees worked, occupy space in office suites in 4 buildings spread throughout a
single suburban office park. This fragmentation of offices spaces was an unplanned
byproduct of the company’s rapid expansion, and the company had plans to build a new
headquarters in order to consolidate its employees. As part of these plans, the company
was, at the time this research was conducted, experimenting with new office
configurations and types of furnishings in its headquarters offices, which played a role
in the research described below. All of the research activities of this dissertation took
place in a single one of these sets of office suites — that housing the technical
communicators as well as the company executives, administrative staff, and a

significant portion of the software developers.

5]

I selected a software company as my research site because the software industry
is prototypical of the type of knowledge-intensive, symbolic analytic workplaces
described by Reich. Further, Software Unlimited’s comparatively long (for the
software industry) history combined with its rapid growth made it a good site in which
to observe a range of issues, infrastructures, and situations in which writers would play
an important role in the memory work of the organization.

In fact, Software Unlimited proved to be an ideal site in which to explore all
three types of memory issues in inventional activities that were outlined in Chapter 1:
responsiveness to kairos, the importance of offloading or distributing information, and
the necessity of personalizing information. First, Software Unlimited’s focus on the
production of software for the consumer market, particularly its expansion into Intemet-
based products, creates a high-pressure work environment of code freezes, shipping
deadlines, and attention to the importance of so-called “Internet time.” These
considerations make responsiveness to kairos a key element in the company’s
employees’ success, thereby creating ideal conditions for the types of scenes of memory
work required by this research design.

An additional impetus for situations of kairos at Software Unlimited is the
company’s adherence to an “Agile” software development methodology. According to
Schwaber and Beedle’s Agile Software Development with Scrum (2002), a volume
which is recommended to all new employees of the company, the Agile methodology
places a heavy emphasis on responsiveness to time. In this, Agile is like the sport of
Rugby from which it draws its terminology: “Both are adaptive, quick, self-organizing,

and have few rests” (Schwaber & Beedle, 2002, p. l).

52

According to Schwaber and Beedle, the Agile methodology borrows from
process control theory the notions that all aspects of the software development process
should be transparent to management and continuously verifiable empirically. The
principal mechanisms enabling this transparency are the Scrum Team and the Sprint.
Scrum teams are small (ideally no more than seven to nine person), multidisciplinary,
and, within limits, self-directed and self-organizing groups of workers who are tasked
with “delivering new executable product functionality” during fixed thirty day design
periods, termed Sprints (Schwaber and Beedle, 2002, p. 9). The goal of each team,
then, is to create a product or piece of a product that could theoretically stand on its own
as an upgrade or stand-alone product. Each Scrum team must contain all the expertise
necessary to complete these deliverables: “Regardless of the team composition, it is
responsible for doing all of the analysis, design, coding, testing, and user
documentation” (Schwaber and Beedle, 2002, p. 37). At Software Unlimited, each
Scrum team that I studied included one technical communicator (titled “information
developer” at this organization), one or two quality assurance (QA) team members, a
Scrum team leader, a user training specialist, and four to six software developers.

The thirty-day Sprints are bracketed by planning and review meetings in which
team members, management, and users determine the functionality to be built by the
team during the upcoming thirty-day period or review the results of the previous thirty-
day period. The most important meetings and mechanisms for transparency prescribed
by the methodology, however, are the Daily Serums. The Daily Scrum is a short
meeting lasting no more 15 minutes in which all team members meet face to face or via

conference call and answer for each other three questions: “What have you done since

53

last Scrum? [. . .] What will you do between now and the next Scrum? [. . .] What got in
your way of doing work?” (Schwaber and Beedle, 2002, p. 43).

The rules, which the methodology urges be strictly followed, are that the
location and time of the meeting be constant, that all team members arrive on time, that
team members speak briefly and to the point when answering the three questions, and
that any design issues or problem discussions be deferred until after the Scrum. The
Daily Scrum, in other words, is meant to be a method for maintaining team and
management awareness, not a working meeting. Although different teams that I
observed adhered to the rules of Serum to varying degrees, one factor that seemed
common to all was the practice of holding short informal follow-up meetings
immediately after the Daily Scrum in which impediments or design issues could be
discussed by smaller subsets of team members. The information developers, in
particular, made frequent use of the Daily Serums as opportunities to arrange such short
face-to-face meetings with developers and other team members.

Each of these elements of the Agile methodology — teams, Sprints, and Daily
Serums — contributed in varying degrees to the conditions for kairotic action in the
research participants I observed at Software Unlimited. The Agile methodology also
contributed to making Software Unlimited a good site for studying the other two aspects
of memory in invention processes. First, the intensive, self-directed, and transparent
nature of the Scrrun teams created a work environment that encouraged distribution and
sharing of information. As just one instance of this, during the Daily Scrum meetings

team members met in spaces where they could view shared Objects of attention: “The

54

team should arrange themselves in a circle, generally around a focus such as a table”
(Schwaber and Beedle, 2002, p. 43).

Finally, the so-called “empirical” aspect of the Agile methodology, with its
concern for observing people and processes and for listening to feedback foregrounded
the role of embodied interactions in meaning making: “Daily Serums provide a direct
view into each team’s progress [. . .] team spirit, each member’s participation, team
member interaction, work that is being completed, decisions that need to be made, and
impediments that need to be removed” (Schwaber and Beedle, 2002, p. 69). Lastly, the
type of software products that Software Unlimited produced (e. g., image manipulation
and screen recording) themselves served to foreground issues of interfaces, interaction,

and embodiment among participants.

The Research Participants

Software Unlimited management helped identify employees for participation
based on their job responsibilities as technical communicators in the Information
Development team. Although each information developer was assigned to a separate
software product and served on that product’s development team, attending Daily
Serums and other Sprint-related functions, management reporting and administrative
functions were performed in the Information Development team by the Information
Development manager, who also participated in the study.

The research participants consisted of four women and two men, each with
varying degrees of experience as technical communicators and different lengths of
tenure with the company, ranging from a few months to almost six years, thus giving

each writer a different perspective on the memories of the organization. The writers

55

chose or were assigned the pseudonyms Angela, Peter, Lance, Monica, Lucy, and

Becky, the team manager.

Table 1: Research Participants

 

Participant

Background and role in the study

 

Angela

Most senior member of the Information Development team with a
tenure of nearly six years with the company at the time of the initial
interview.

Participated in the initial interview and meeting observation session.

 

Becky

Information Development team manager for just over one year at the
time of the initial interview. Worked as a member of the Information
Development team for six months before becoming manager.

Participated in the initial interview.

 

Lance

Newest member of the Information Development team with a tenure of
only a few months at the time of the initial interview.

Participated in the initial interview, composing and meeting
observation sessions, and follow-up interviews.

 

Lucy

Employed by the company for two years at the time of the initial
interview, though had only recently returned to full-time work after
time off and part-time work during pregnancy and an illness.

Participated in the initial interview, composing and meeting
observation sessions, and follow-up interviews.

 

Monica

Three years with the company. Started as an intern while in college.

Participated in the initial interview and meeting observation session.

 

Peter

 

 

Just over one year at the company at the time of the initial interview.

Participated in the initial interview.

 

Physical and Digital Infrastructures

As mentioned above, the Information Development team was located in a single

suite of offices, which they shared with several other groups of employees (see Figure

2). Peter, Lance, and Lucy were assigned to cubicles in the center of a large room in

one wing of this suite, and Monica, Angela, and Becky occupied offices around the

perimeter of this room. Other company employees (i.e., employees who were not

56

 

information developers) occupied the remainder of the cubicles and Offices in this room.
One end of the large room was devoid of cubicles and left open in order to be used
exclusively by the product team to which Angela belonged. This open area and the
offices immediately surrounding it were nicknamed the “Petri Dish” because the area
contained experimental furnishings and office configuration for the planned new
corporate headquarters. The furnishings in the open area, including tables, chairs, white
boards, and a video projection screen were all wheeled so that they could be

reconfigured as needed. Likewise, the offices at this end of the room held two

Figure 2: Information development wing

 

 

 

 

Ira

 

l . r ‘
.. ... _“ “T —h 4' T Scrum
O Becky Monica Room

,4-.. ID A6 Rese:l:er _i® TL
D (2

 

 

 

 

 

 

 

"Petri Dish"
Area

 

 

 

 

 

¥
6

 

 

 

 

 

 

 

 

 

 

 

 

4— Angela

 

 

 

. W
iMRIZRr *

employees each. These employees’ workstations within the offices were placed in

 

 

 

 

 

 

several experimental configurations -— some occupants’ desks faced each other, others

faced away from each other, and still others faced the same direction. The information

57

developer Angela occupied this last type of experimental configuration. Two other
Information Development team members, Monica and Becky, the manager, occupied
conventionally-arranged and fumished offices along the perimeter of the other end of
the room. Lance, Peter, and Lucy occupied separate cubicles in the middle of the room.

Their cubicle walls were five feet tall.

At one end of the large room, there were two meeting spaces: the “Library” and
a small meeting room termed the “Scrum Room.” The Library held a large conference
table with seating room for ten, a projection screen, and bookshelves along the
perimeter, filled with books and periodicals of importance to the software industry, as
well as video games and movies which could be checked out by employees. The Scrum
Room possessed about 10 plastic chairs and a single work table, which remained pushed
to one side during the period of the study. Two walls contained white boards, one wall
a large window, and the other a projection screen. The projector sat on the table.
Although some differences in each writer’s workspace existed and will be noted later
where necessary, each workspace possessed at a minimum a desk, chair, small file

cabinet, laptop or desktop computer, and at least one monitor.

Most of the data collecting reported in this dissertation took place in the spaces
described above, including all of the composing observation sessions and several of the
interviews. The interviews with Lance and Lucy, however, took place in a smaller
conference room in another part of the building. Also, Lance’s meetings with
developers took place in the developer’s offices or in the hallway where his Daily

Serums were held.

58

As would be expected at a software company such as Software Unlimited, the
computer hardware and software infrastructures supporting the information developers
were extensive and participants’ configurations and use of these infrastructures varied,
as will be noted in the analysis chapters. However, a few commonalities across
participants can be given here. Every participant had at least one personal computer in
his or her workspace, though several participants had more than one, including a mix of
laptops and desktops and Macs and PCs. Every participant had at least one external
monitor and keyboard, though, again, many had more than one. Every participant had
the Microsoft Office suite installed on his or her computer, as well as Software
Unlimited’s own suite of graphic editing software. Every participant used Microsoft
Messenger for instant messaging and Microsoft Outlook for email. Additionally,
several participants used the Outlook calendar feature for work scheduling and
reminders. Every participant had access to the Internet, a shared network drive, a

shared company intranet, and a shared company Wiki.

Data Collection Methods and Rationale

Studying memory work as scenes occurring at the confluences of rhetorical
purposes and bodies and tools in space and time required a mix of data collection
methods. That is, in order to assemble a scene by indexing observable activities to the
rhetorical purposes and situational exigencies for those activities, it was necessary to
collect as much data from as many sources as possible. Only by employing a mix of
ethnographic methods including observations, videotaping, and interviewing was it

possible to render visible the operations of memory as enacted in embodied activities.

59

Using such a mix of data collection methods also helped triangulate findings
(Lauer and Asher, 1988; Denzin & Lincoln, 2003; Krarnp, 2004; Hays, 2004). My
purpose in such triangulation was not in order to claim objective proof of the
phenomena I observed, but instead to ensure that my accounts of the scenes of memory
recounted in Chapters 4, 5, and six possessed verisimilitude. Krarnp (2004) defines
verisimilitude as “the appearance or likelihood that something is or could be true or
real,” which is possibly as close to proof of the fleeting and ephemeral functioning of

memory in activity as we are likely to get (p. 108).

The methods described in this section enabled me to assemble the scenes of
memory work by providing me with data about what counts as stored information or
memory in this organization, where this information lives, how different types of
memories are valued, and how these memories are stored and retrieved in practice. The
methods also helped me to answer, at least tentatively, the research questions detailed at
the end of Chapter 2 concerning how, when, and where writers use this information
when composing. Table 2 maps each research method and the type of data the method
generated to the element of the scene of memory work that the method helped explicate
and to the research questions it helped to answer. Following the table, each research
method is explained in more detail, including its rationale for inclusion in this study.

Finally, the data collection schedule is summarized in a table at the end of this section.

60

Table 2: Data Collection Methods Mapped to Scenes and Research Questions

 

Data collection

Element of scene this data

Research question it

 

method and data helped reveal helped answer

source

Dwelling (field Rhetorical purposes and Where/when are embodied
journals) situational exigencies contexts most important for

memory work?

 

Initial interviews
(video recordings)

Rhetorical purposes and
situational exigencies

Do writers use the
affordances of their
embodied contexts to aid
memory work?

 

Composing session
observations (field
journals and Morae
recordings)

Memory activities

Where/when are embodied
contexts most important for
memory work?

How do embodied contexts
help?

 

Meeting observations
(field journals and
Morae recordings)

Memory activities

Where/when are embodied
contexts most important for
memory work?

How do embodied contexts
help?

 

Follow-up interviews
(field journals and
Morae recordings)

Rhetorical purposes and
situational exigencies

Where/when are embodied
contexts most important for
memory work?

How do embodied contexts
help?

 

 

Artifact collection
(soft and hard copies
of documents)

 

Memory activities

 

Where/when are embodied
contexts most important for
memory work?

 

Dwelling

As part of a case study rationale seeking to provide “in-depth descriptions and

interpretations,” I spent around 120 hours over a six month period (September 2007

through March 2008) at the research site in activities that I label “dwelling” (Hays,

61

 

2004, p. 218). As part of these dwelling activities, beginning during my first week at
the research site and continuing over the course of the entire six months of the study, I
attended all but three of the weekly Information Development team meetings, sat in on
a dozen meetings between the information developers and members of other teams, and
spent ten entire days working, listening, and watching in a cubicle that Software
Unlimited had provided for me adjacent to the information developers’ workspaces (see
Figure 2). My purposes in these activities were to help participants grow comfortable
with my presence in their midst, to get a sense of the daily flow of activities, and to
attempt to learn about job functions and team relations so that the subsequent interviews
could be more focused. My own experience as a technical communicator, including
four years at a similarly-sized software company, contributed positively to my comfort
level and, I believe, positively influenced my ethos with my participants. On several
occasions I was asked about my experiences and I was also asked to contribute to

several team-building activities.

Initial Interviews

Beginning at the end of the first month of the study, I interviewed each
participant to learn about his or her background (including educational background,
history with the company, and experience as a technical communicator), and memory
and writing tools and practices. In Yin’s (2003) taxonomy of interview types, these
interviews were focused but open-ended. That is, they were relatively short — lasting
from 45 minutes to one hour - and consisted of the same core set of questions for each
participant (see Appendix), but I conducted them in an open—ended manner by adopting

a conversational tone, by asking unscripted follow-up questions where I thought more

62

information would be helpful, by omitting certain questions if I thought the answers
would be redundant, and by interjecting anecdotes from my own experience where I felt

they would be appropriate.

These interviews were audio and videotaped using a Webcam attached to my
laptop. The Webcam was pointed at the participants throughout the duration of the
interview in order to record participants’ gestures and movement as they answered the
interview questions. The purpose of videotaping the interviews was to provide
additional data sources in order to create a richer understanding of the unfolding dialog.
In this, I was guided by the work of Leander & Prior (2004) who argue for the
importance of capturing data about embodied practices when studying composing by
asserting, “speaking (and silence), gesturing and the text work together in [. . .] meaning-

making; it is impossible to interpret meaning form any one of these modalities alone”

(p. 231).

An even more important reason for videotaping was the recognition that the
interviews themselves constituted scenes of memory work in which participants used
embodied resources to facilitate recall and to help them compose answers to my
questions. The insight that interviews themselves are key sites of memory work is
offered by Yin (2003), who although he is making rather a different point, makes an
important observation about the difficulty of doing memory during interviews,
“interviews should always be considered verbal reports only. As such, they are subject
to the common problems of bias, poor recall, and poor or inaccurate articulation.” (p.
92). In short, I believed that the interviews would be a good place to look for some of

the same embodied articulation work as any other composing activity.

63

Composing Session Observations

Two of the writers, Lance and Lucy, participated in work-session observations
in order to closely observe the moment-by-moment composing activities that occur in a
scene of memory work. These four sessions are the primary focus of my analysis in
Chapters 4 and 5. I selected Lance and Lucy because of their high level of comfort with
the research process, which, I reasoned, would lead to more natural and therefore more
valid data. I employed two field-work methods to capture data during these observation
sessions. First, in the two observation sessions with Lance, I sat behind him in his
workspace and took handwritten field-notes as he composed and, afterwards, as he met
with software developers. During these sessions, Lance would provide clarifying
information about his work processes whenever I asked or, more frequently, when he
felt additional information was necessary; however I tried to keep my questions to a

minimum in order to not impact Lance’s usual work processes.

The second field-work method, which I employed during two composing
sessions with Lucy, utilized the user-testing software TechSmith Morae to record
Lucy’s on—screen activity, her interactions with her computer (mouse and keyboard
interactions), and her physical gestures and movements (via Webcam placed on top of
her computer monitor). Morae also enabled me to watch remotely from another office
in real-time and to take field—notes as Lucy worked during these sessions. Lucy already
had a Webcam and Morae installed on her computer and routinely used Morae to record
herself working for the purposes of usability testing, so the recording process did not

seem to present a major distraction.

64

Each of these work sessions attempted to capture a naturally-bounded segment
of working time, lasting anywhere from 45 minutes to an hour. In the two sessions with
Lance, the Daily Scrum meetings provided the terminus for the composing session, and
provided the direct exigency for the composing task as he scrambled to prepare
materials for the Scrum or post-Scrum meetings. The two sessions with Lucy occurred
at natural workday boundaries: the first when Lucy stopped work for the day in

preparation for going home and the second at the end of a specific work task.

Meeting Observations

When I began this study, it quickly became apparent that one of the principal
mechanisms through which invention at Software Unlimited takes place is meetings.
This is unsurprising since the importance of meetings to the conduct of symbolic-
analytic work has long been recognized by researchers of knowledge-intensive
workplaces (Brown & Duguid, 2002; Winsor, 2003; Johnson-Eilola, 2005). For the
Information Developers at Software Unlimited, however, the Daily Serum and the
smaller post-Scrum meetings appeared to be of particular importance to their inventing
and composing. Consequently, this study reports the findings from observations of two
Daily Serum and post-Serum meetings between Lance and several software developers.
Data from the first of these meeting observations was recorded entirely via my
handwritten field-notes, while the other was recorded with field-notes and videotaped
using a Webcam placed so as to capture the speech and gestures of the participants. 1

used this video to check and clarify my field-notes when I wrote them up.

65

F allow-up Interviews
As soon as possible following those observation sessions in which I was not free

to ask clarification questions of participants as they worked (i.e., meeting observations
and sessions observed remotely using Morae), I conducted debriefing session
interviews with participants. These short, informal interviews were loosely structured
based on questions I had noted in my field-notes or using the Observer Log feature of
Morae. The purpose of these interviews was to better understand any actions or other
elements of the scene that were not readily apparent to me based on the context. An
example of this type of question included asking Lucy to explain the purpose behind a

highlighted portion of her text that I observed during one composing session.

Artifact Collection for Corroboration

Whenever possible, after composing or meeting observations, I requested hard-
or soft-copies of any texts the participants interacted with during the session. Instances
of these artifacts included participants’ texts-in-progress, their to-do lists, and their
meeting notes. Having these artifacts provided a fuller picture of participants’ memory
practices in given scenes. For instance, several texts-in-progress contained lacunae
where participants were unable to find the information needed to complete the
document, while others served dual purposes as both talking-point reminders and as

objects of joint attention during meetings with software developers.

Data Collection Schedule

Altogether I logged over 120 hours at the research site between September 2007

and March 2008. Table 1 gives the dates of the interviews and observation sessions.

66

Table 3: Data Collection Sessions

 

 

 

 

 

 

 

 

 

 

 

 

 

Participant Initial Composing Meeting Observation
Interview Session
Observation
Becky 9/19/07 —- --
Lance 9/26/07 10/25/07 10/25/07
11/21/07 11/21/07
2/6/08
Peter 1 0/4/07 -- --
Monica 10/18/07 -- 8/29/07
Angela 10/24/07 -- 9/13/07
Lucy 11/16/07 1/17/08 1/17/08
2/29/08
Data Preparation

My research methods resulted in four basic categories of data: 1) the
handwritten field-notes that I took while observing a large number of meetings, three
composing sessions with Lance, and follow-up interviews with Lucy; 2) videotapes of
initial interviews with all five participants and a meeting between Lance and a Web
developer; and 3) Morae video/screen recordings of two composing sessions with Lucy;
and 4) artifacts. Each of these data types required different steps in order to prepare it
for segmentation and coding. For handwritten field-notes, as soon as possible after an
observation session (typically within 24 hours), I converted my notes into coherent
write-ups and attached to these write-ups cover page contact summary forms (Miles &
Huberrnan, 1994) on which I recorded salient contextual information about the session,
including the participants involved, the location, any software or hardware employed by

the participants, and any particular issues, themes, or unanswered questions related to

67

my research questions that arose during the scene. The composing session with Lance
on 2/6/08 had to be removed from the data analysis reported in Chapter 4 because of my

failure to produce one of these write-ups in a timely fashion after the session.

For Morae recordings of initial interviews and Lance’s meeting with the Web
developer, I transcribed the audio and made extensive annotations from the videos,
noting participant gestures, facial expressions, posture, and object manipulations
indexed to the corresponding audio. The two work sessions with Lucy recorded with
Morae captured her onscreen actions (mouse cursor movements, clicks, and keyboard
activities), audio, and video from a video camera placed atop her monitor pointing
directly at her face and automatically indexed to the corresponding onscreen actions.
For these recordings, I transcribed speech and wrote detailed accounts of Lucy’s
activities on her computer and any gestures or physical object manipulations captured
on the video. Additionally, as with the work sessions recorded using field-notes, as
soon as possible after the observation session, I created cover page contact summary

forms on which I recorded background information about the session.

Data Segmentation and Coding

I segmented transcriptions of interviews and meetings into discourse units or “d-
units” (Colomb & Williams, 1985). Colomb and Williams (1985) define a d-unit as
“any stretch of continuous text — a whole text, a section, a paragraph, even a small
group of related sentences — that functions as a unit and whose parts are more related to
each other than to those outside the d—unit” (p. 102). They further elaborate that the d-
unit is an appropriate tool for rhetorical analysis because it helps us account for

contextual features of connected text that other analytic tools, such as the t-unit, do not:

68

“what we lack is a consistent, systematic vocabulary that correlates perceptions of

coherence with features of text structures” (Colomb & Williams,1985, p. 101).

Although Colomb and Williams first formulated the concept of a d-unit as a
means of describing structure in written rather than spoken texts like interview and
conversation transcripts, I selected it as my unit of analysis because the d-unit, more
than any other analytic unit, met my need for understanding speech about activities as
indicators of those activities. The central structure of a d-unit is an issue statement and
a discussion about that issue. Therefore, I reasoned that the d-unit, as a coherence-
explaining structure, corresponds well with the question and answer format of
interviews in which I asked participants to describe the work processes that I later
hoped to see in action: my questions demarked an issue and the participant’s answer to
the question constituted the discussion. In other words, I believe that, as it does with
written text, the d-unit possesses descriptive power for the coherence that interlocutors
perceive and achieve during structured conversations such as interviews as the
interviewee listens to and then responds to the interviewer. In short, if we substitute the
word “interlocutors” for “readers” in the following passage, Colomb and Williams’
(1985) point applies equally as well to structured conversations as it does to written
texts: “coherence is not an inherent feature of texts. Readers create coherence and, for
the most part, are eager to do so. They will do so when the texts they encounter allow

them to generate appropriate expectations for what is to follow” (p. 105).

Two further aspects of the d-unit, as Colomb and Williams formulate it, are also
useful. First, the component parts of the d-unit, the issue and discussion, do not

correspond to sentences or even groups of sentences, as in the traditional notion of a

69

topic sentence, but instead are “fixed discourse position[s] or slot[s]” that can vary in
length, ranging from a clause to a paragraph, depending on the phenomena of interest to
the researcher (Colomb & Williams, 1985, p. 108). I found this helpful in several
instances for understanding my participants’ answers to my questions (i.e., the
discussion portion of the d-unit) as divisible into multiple d-units. So, for example,
when responding to my interview question: “What resources do you typically consult
when conducting research when you begin a new proj ect?”, Lance began by referencing

other people as resources:

I try to make friends with all the programmers who are doing the website or
doing the recorder or anything. Now it’s actually getting better - on Monday
morning or Tuesday — [Becky] and I had a meeting with some marketing people,
which doesn’t really happen that often. So I’m learning how to get something
done. So today for example, there’s some technical things that I don’t really
understand or spend my time on like how to download or install or point my
recorder to the right server. So now I know this [. . .] guy who’ll take care of that

problem really quick for me.
In answer to the same question, however, Lance continued by talking about a finding
activity:

Other things, [Software Unlimited] on the N drive has a history of all the things
associated with a project. So if I went into [product name] when I got here I
could kind of go back and look at what had been said in other meetings and

Sprints. So far it seems like it’s not really organized or used very well, so one of

70

my goals with the Wiki would be that I could click on [product name] and

there’d be a lot better organization of what’s going on.

The affordances of the d-unit enabled me to segment the two passages separately.

Second, d-units themselves can be nested inside other d-units, again depending
on the phenomena of interest, a characteristic that makes the d-unit powerful in
describing the format of semi-structured interviews and of speech about activities as
indicators of those activities. That is, the concept of nested d-units can help us account
for the structure of the inevitable side-roads and follow-up questions that occur in semi-
structured interviews, but it can also help index talk about activity to the activity such
talk describes. So, continuing with the example above, Lance resruned talk of
referencing once he finished his digression about finding: “If I got transferred to
[another product] that would be really scary to me because I don’t know much about it.
[. . .] So the first thing I would do is talk to Peter who’s on it now and get the behind the
scenes kind of [stuff].” The affordances of the d-unit, then, enabled me to, where
necessary, segment these based on the different memory activities or practices Lance

described.

As for coding these activities, I coded data according to the type of memory
activity being performed (in the work session observations) or indexed (in the
interviews). This coding scheme is based on my reading of the vast and
multidisciplinary literature on memory. However, because this literature was so vast

and multidisciplinary, arriving at definitions and boundary conditions for the activities

71

of “doing memory,” of necessity, involved a degree of interpreting and eliding to better

fit my focus on memory as rhetorical action, as I will explain below.

The codes that I adopted were archiving, reminding, finding, referencing, and
gesturing. First, I labeled as archiving any activity or practice whose primary purpose
was storing knowledge for later use. I chose the term “archiving” over other
possibilities like “storing” because of the richness of the term in the literature, including
especially Derrida’s (1998) emphasis on the jussive and sequential aspects of archiving
and Bowker’s (2005) focus on archiving as the process of rendering knowledge useful
through classification: “a structure of record keeping [that] subtend[s] this common
time, rendering it useful through permitting the collocation of accounts of said events”
(p. 10). The following are examples of activities that I label archiving: taking notes at a
meeting; creating a directory structure on a hard drive; storing a document in a paper

file folder; copying a file to a network drive.

By contrast with archiving, I use the term reminding as shorthand code for an
activity which could best be thought of as “creating reminders.” Where archiving
activities entail saving memories for the long term and for multiple or not always
clearly-defined future uses, I distinguish reminding as creating a limited-term memory
(basically a memory aid) with a single clearly-defined purpose or a very limited number
of purposes. Where archiving primarily facilitates conscious searching and browsing,
reminding principally facilitates unconscious recognizing. Larsen and Wactlar (2003)
distinguish recognizing from finding, which they term recalling, this way: “cognitive
psychologists distinguish between two fundamental types of memory: recognition and

recall. Recognition occurs when you see something familiar, while recall requires that

72

you remember something and are able to articulate it” (p. 15). The following are
examples of activities that I label reminding: creating a list of questions to ask at a
meeting; creating a post-it note and sticking it in a spot where it will be seen; using the

calendar feature of Microsoft Outlook to create a pop-up meeting reminder.

1 code as finding any activity in which a participant attempts to locate digital or
non-digital memories. For finding performed with computers, this principally includes
the activities of conventional information retrieval: logical searching and location-based
browsing, to use Barreau and Nardi’s (1995) language. According to Larsen and
Wactlar (2003), in logical searching “the user has to describe what he/she wishes to
retrieve, formulate a query, and submit a search [while in] browsing [you have to look]
around until you find something of interest that you recognize as useful [. . .] most
browsing still requires that the user describe a starting point.” (p. 15). In sum, I
operationalize logical searching as a keyboard-enabled activity in which a user types
text that he or she wishes to search for and location-based browsing as a mouse-enabled
activity in which a user navigates a graphical user interface of some type, such as
Windows Explorer or the menu structure of a software application. I also attempt to
take into account intentionality: Where recognizing a reminder is often unintentional,
finding something stored in an archive is intentional (searching) or semi-intentional
(browsing). The following are examples of activities that I label finding: browsing
through files on a hard drive; querying using the Find and Replace feature of Microsoft
Word; nunmaging around an office; searching the Web; looking for reading material in

the company library.

73

Ackerman and Halverson (1998) point out that organizational memories are
“complexly distributed, interwoven, and occasionally overlaid. They [are] sometimes
the province of the individual [. . .] or the group” (p. 46). Distributed cognition theory
recognizes that memories are also distributed to artifacts (Hutchins, 1995b). So, I code
as referencing, then, activities in which the knowledge of another person or artifact is
consulted or referenced in speech. In a sense, then, referencing could be thought of as
the activity that follows successful finding: after we have found a source document, we
reference it as we write as an aid to short-term memory, or we ask a colleague for help
on a particular problem after we have found out who possesses that knowledge and how
to go about asking for it. The following are examples of activities that I label
referencing: reading information from a white board or a document; asking someone for
information during composing; referring to knowledge or a type of knowledge held by
another person during an interview; alt-tabbing to look at the user interface of an

application while writing documentation for that application.

Finally, borrowing from the gestural coding method advocated by McNeill
(2005), if the participant performed a gesture high in the dimensions of iconicity
(picture drawing) or deixis (pointing), I double-coded the segment gesturing. My
rationale for paying attention to these types of imagistic gesturers is that they represent
an important means of “seeing” what is going in my participants memories as they
speak and work. According to McNeill (1992), “gestures are the person’s memories
and thoughts rendered visible. Gestures are like thoughts themselves. They belong, not
to the outside world, but to the inside one of memory, thought, and mental images” (p.

12). Further, gestures give a glimpse into the spatial, embodied nature of a person’s

74

memory: “The gesture reveals not only the speaker’s memory image but also the
particular point of view that he had taken toward it” (McNeill, 1992, p. 13). In other
words, I believe that gesturing, particularly gesturing that reflects my participants’
memory-images of scenes, plays an important role in helping me understand the role of

embodiment in memory.

Departing from traditional methods of coding gestures such as McNeill’s (1992,
2005) schemes in which he labels as “gesture” only those physical movements
coinciding with speech, I also code as a form of gesturing mouse cursor movements that
do not entail software manipulation, which, for lack of a better term, I label “touching.”
In more specific terms, these are mouse cursor movements that do not play a role in
triggering “events,” such as clicking a button, scrolling a menu bar, highlighting a text
in order to copy it, hovering the cursor over hyperlinked text in order to trigger a pop-
up, or dragging a dialog box. Instead, they appear to be some form of deixis, in which
participants attempt to physically touch the information displayed on their screens. As
with hand gestures, then, I am interested in these unintentional or semi-intentional

gestures as indicators of embodiment in memory work.

Data Analysis

The purpose of my analysis of the scenes from Lance, Lucy, and Angela’s
working lives was to understand the role of embodied contexts in mediating the
memory-related activities of invention. As my research questions state, I wanted to
know if these contexts played a role; if so, when and where they were most important;
and, finally, how they helped or hindered composing. The process of analysis, then,

involved searching for patterns of the categories of memory work identified above in

75

these scenes, and contextualizing and corroborating those patterns with reference to my
entire corpus of data from all my participants’ meetings, job shadowing, and interviews.
My method of presenting this research in Chapters 4 and 5 is to provide an introduction
to the participant (Lance in Chapter 4 and Lucy in Chapter 5), an overview and synopsis
of the two scenes in which I attempt to highlight the action of the scenes that I found
most salient to my purposes, and finally, an analysis in which I offer tentative
hypotheses of the patterns I found in each of their scenes, again corroborated by data
taken from the entire corpus. Finally, Chapter 6 attempts to bring these patterns and
hypotheses together by considering a scene of memory work with Angela in order to
offer some tentative answers to my research questions and to suggest future directions
for research into the embodied theory of memory in invention that I am attempting to

formulate.

76

CHAPTER 4: TWO SCENES OF MEMORY WORK WITH LANCE

Introduction
In this chapter, I focus on Lance’s memory activities in two scenes of memory
work that I observed. I will argue that there is a pattern to the ways in which Lance
engages in these activities and that the other information developers display a similar
pattern, and hence that the activities constitute a practice. Moreover, I will argue that
this practice of memory is shaped in large part by the memory regime at Software

Unlimited.

My method of analysis is to compare two of the memory activities from my
coding scheme — archiving and reminding - as a way of speaking to a range of memory
practices in this regime. It is, therefore, worth repeating here that the principal
distinction between archiving and reminding as I define it lies in differences of time and
purpose: archiving practices have multiple, often not clearly-defined purposes and are
intended to store memories for the longer-term while reminding practices are generally
shorter term and have a single (or quite limited) number of purposes. As the scenes
recounted below show, the distinction between these “input” activities of memory is not
always easy to discern and one activity can often merge into the other, but, as I hope my

analysis reveals, it is precisely these overlaps that can provide the most insight.

At the time of our interview and Observation sessions, Lance had been employed
at Software Unlimited for less than half a year, making him the newest member of the
Information Development team. Despite this, he came to the company with

considerable previous experience working with technology and had already earned the

77

respect of his teammates, who highly praised his knowledge and research skills, one
teammate averring “he’s really great at research. We haven’t worked together that long
but he comes up with stuff [. . .] I want to learn more about how he does research in the
future because he seems to come up with a lot and come up with new ideas from places

I never would have thought to look.”

This high degree of competence and desire to learn was demonstrated in my
interview with Lance, in which he informed me of his enthusiasm for working for a
company that values innovation and encourages its Information Development team to
act as user advocates (a duty that was actually listed in the job description) by helping
shape the design of the product user interfaces and workflows. In fact, all five
information developers, including Lance, informed me that acting as user advocate in
the software design process was their primary job duty, more important even than
writing documentation, one participant even adding that “writing per se [is] the smallest
part of our job.” Not surprisingly, then, persuading the software developers to improve
the user interface workflows was the substantial motivation for both of Lance’s
composing activities described in the scenes in this chapter. The company’s
encouragement of innovation also influences Lance’s archival practices. For instance,
he keeps a directory on his computer titled the “back of the mind” because “it’s
conceivable that some of those crazy ideas that I might get a chance to do some of them,
so it’s worth writing them down whereas at some companies it’s like don’t even bother

thinking about them.”

Since he arrived at the company, Lance has been assigned as the primary

information developer on a new product, a “lite” web-based version of an older

78

conventional “boxed” (i.e., sold as a disk in a box) Software Unlimited product.
Understanding and influencing the evolving workflows of this new software has
become Lance’s primary preoccupation and it influenced many of his job activities and
his attitudes toward these activities that I witnessed during my research at the site. For
instance, Lance interpreted my question asking him to describe his project management

practices entirely in terms of understanding his product. Of the product he says:

Even though it doesn’t exist yet, I pretend it does and I might use freehand paper
and I draw out what I think it might look like even if the UI people don’t come
up with the design or anything. And then I try to write help for it even before it
exists basically. That really helps me get the workflow stuff down and then [. . .]
we might see a mockup or something and I see if I compare my workflow to that

document how would it work.

This quotation points to an important domain of knowledge and means of storing that
knowledge in this organization: the product itself. I will be examining this in more

detail in the analysis which follows.

The scenes recounted below took place during job shadowing sessions in
October and November 2007. I selected these days and times for job shadowing and for
more detailed work session observations for the following reasons in order of priority:
1) Lance’s indication that these would be convenient times for me to observe; 2) his
consideration of the sessions’ relevancy to my project (i.e., he knew that he would be

engaging in extended composing tasks within bounded spans of time. Both composing

79

sessions were terminated by his daily Serums, which I was also allowed to observe and

which also formed part of the scene); and 3) my own teaching schedule.

Scene 1: Arguing for Changes to the User Profiles

Overview of the Scene
This scene took place the morning of October 25, 2007. The activities of this scene of
memory work occur in three parts, each with a different cast and setting. The first part
of the scene took place in Lance’s cubicle from about 9:00 am until 10:30 am, and
primarily involved Lance working alone on a documentation task while I sat behind and
slightly to the right of him, observing and taking field-notes (see Figure 3). He had his
laptop open on his desk. This laptop was attached to an external monitor, keyboard, and
mouse, and sat slightly to his right while the external monitor sat directly in front of
him. He had enabled a function that allowed his laptop display to act as a subsidiary
display for the monitor so that he could view information on both. The second part, the
Scrum meeting, took place from 10:30 am until 10:40 am in the “Scrum Hall” and
consisted of the 12 product team members standing in a semi-circle facing a large Gantt
chart (see Figure 4) detailing each team-members’ progress toward the 30 day Sprint
goal. The final part of the scene began immediately following the Scrum, when Lance
asked two developers to stay behind and meet with him. The three remained standing in
the Scrum Hall for the duration of this short meeting (about 10 minutes). In both the
Serum and the follow-up meeting, I stood slightly behind the participants in order to see

and hear the discussion without being obtrusive.

80

Figure 3: First part of scene one - Lance’s office

 

Laptop, external
/ monitor, mouse,
and keyboard

/

 

 

Lance

 

 

 

Researcher

Figure 4: Second part of scene one — Scrum hall

 

81

Synopsis of the Action of the Scene

When I arrived at 9:00 am, I found Lance already at work in his cubicle. He
informed me that his main task for the period I would be observing was to start updating
a section of the draft of the product user guide to reflect changes in the user interface
and in the workflow for updating user profiles in his product. He further informed me
that he was troubled by the latest version of the user interface, which had been changed
completely from earlier iterations, and that he believed it “wouldn’t work” because it
would be too complex for most users to update. He explained that this was particularly
frustrating because a big goal of his product team all along had been to make setting up
profiles easy. He also added that the “End of Sprint demo” of the latest version of the
product to the whole company was scheduled for November 6, so any changes to the UI
would have to be completed no later than November 1, six days from the time of my

observation session.

Lance’s first action as he began working on this task was to open the most
recently updated draft of the user guide and to print a copy; however, rather than
working from this draft, he chose to start a new Word document from scratch. He
informed me that this was because the extent of the changes to the product had made
him decide that it was better to start a new document than to update the old, even
though the deadline was looming. He created the primary headings in this new
document, naming them “Set up users,” “Create a new profile,” and “Assign profiles.”
Noticing that he was not referencing the printed version of his draft as he wrote these
headings, I asked him if he was working from memory. He informed me that he was

“thinking like an admin,” or thinking about the prototypical administrator tasks that

82

would need to be performed in the product. Thinking and acting in the person of the

user was to be a recurring theme with Lance in my observations.

As he began to write the text for the “create a new profile” section (again not
looking at his draft), Lance extensively referenced his product, which was running on
his secondary laptop, but he did not appear to refer to the original draft or to the printout
very often. He worked on this new document for about 20 minutes, alternating his
attention between the document-in-progress and the product several times, and quickly
writing about one page of documentation (see Figure 5) as he navigated through the

product. At about this time, Lance turned to me and told me that he was beginning

Figure 5: New section of the user guide

 

Create a New profile
Y: mm s;- a -::.r,-.:L.r. us: 2.- '.: -::...r !$:".\:.:.‘_.".1'..:‘. l.)
‘. (J- ‘- lr-fJ. s.
: (4— :w «A: ::r
. . ¢

 

’31 .1 : ‘.n;; :.'.n -‘.r.u..“ ‘.I ;.!,I.‘t...'.0 LI 35:51.: Darn. ‘ r..'.l
$3.4}: u.:. ;" l L: Am I ..IL:;- -_r: try. :;a.-‘:: "s- .':. 1.214 F.:’.1
1: :".'v :. pr". 1.“..3 Kraft. r. 1:12;": '.. i‘.

.; .0'. I" .c.~:'.o 0"Ils:s

Jmcr‘ro met"; fierce r 59.: '6' “my

7.2: am: 12.3.: .l‘. :4 .f '12“: in": 1'“. reflux; :4: 52.3-;
. Tern-Ir 'r ‘(:x.«‘.". 4.“ 1"\’A‘.Q -'- AJ.O'| r4“- o;a’. ahgsvt
: c.“ n...“ moxsc1°gte~ new: 0' ' an Veranda-v
2 :35:r..-1<o:a' . ‘04” 51 lfOKA'L'j "Latin: .

 

\ C r- -l n if .‘ ls--r.;a..r;: 4.. :Sr. .72"; 5: ';:;F:‘e.'-: f: '-.r :1

$1.! 221.“. £51,132! ILL?3.13.3:;‘:-.-,l.r:l‘.:‘.!‘L:.'.!f-:‘-.:‘.:'.-

actual-.1» ‘.r.r.n; ‘;:‘..‘ r.‘ -. .x 11".: a;
" v ..‘..'-’. “sunny“ .1- ' I o¢l -«c.ll:1.‘.'- .-.o.‘
t-ru-';c;a'.r;jru.

: C1,, :'1 .o-a. ‘1\ I‘LV“G'¢OCE 4:1. .- "QJ‘I‘C rut-la" tn! 1 1:20
"a ':".s'o.° l".l rots: ‘5' «one:

I lTAC1s-N(Oi'u" a 1.." A" "Q‘IJ‘T."T nrlrunu '

' (J-lr-r

: (4» 44 l:.."H'.‘|..$‘ul‘t'fb{'.a!'.‘; “raft-.44::.¢‘.'r¢u:‘r31
.:.' . N

O c (u 1" :un. «\na.45;:.-‘.r..1..rz'raza'ch-H'! f“ . ‘0."[0
"'x 9:09,‘:.“.'[<v.'.'ru'a:rtr 10:"

.: Intuwcnu- . 'v..v¢ r-r 3x..-'.'|¢<.'.'su' 11.25!!le ‘ .

.‘. CJ'IL"

:: tx-ls’!

 

 

 

to perceive some serious problems in the product workflow. At this point, he stopped
documenting and began exclusively navigating and experimenting with the product

interface. He had. an intent look on his face as he did this.

83

Lance next informed me that he had already had a conversation with the
software developer about the workflow for this procedure but that he could not
remember exactly what she had told him about it, so he was not entirely sure if the
software developers were planning to change the interface again or not. So, at this point
he created another Word document and wrote down a particular question about the
workflow for the developer, informing me that he would ask the developer this question
after the day’s Scrum meeting. After writing his question, he resumed documenting the
procedure and stepping through the product interface. However, after just a minute, he
realized that he had another question for the software developer and then another and
another and another. He updated his second document, which I am terming his
“reminder document,” with these questions as he worked, alt-tabbing between the
product interface and the two docrunents many times. After about 30 minutes of this,
he opened a screen capture program and began taking screenshots from the user
interface to clarify his questions in the reminder document, even going so far as to add

annotations and callouts to this document (see Figure 6).

While working, Lance wrote himself a post-it note reminder to ask his colleague
Peter to show him the documentation for his own product. Lance informed me that he
likes the user interface of Peter’s product, which he believes “makes the doc almost
unnecessary” and would like to see a similar solution for his own product and
documentation. He further added, that, based on the problems with the workflow and
the number of questions he had for the developer, he suspected that the procedure he

was documenting for the user manual was going to change substantially.

84

Figure 6: Lance’s reminder document

 

 

 

 

 

 

 

 

At 10:25 am, an Outlook Calendar reminder popped up on his screen, giving
him five minutes’ notice that his Serum was about to start. He printed his reminder
document and then saved an electronic copy to his desktop, where he informed me he
tends to save documents that have only a limited lifespan, and saved the new draft of
the user guide to the network drive. He grabbed a pen and his printout of his reminder
document and we hurried off to the Scrum, which was located on the other side of the
building, so that we would not be late (Scrum rules discourage tardiness). As we
hurried, Lance informed me that he uses Outlook reminders because he has trouble
remembering to attend his Serums now that they are no longer held in the Scrum Room

adjacent to his cubicle.

The rules of Serum, which dictate that the Scrum not become a working
meeting, prevented Lance from addressing his issues with the developer in the Scrum
itself, so he simply noted the workflow as an impediment and asked if the developer,

“Andie,” had time to meet with him afterwards. Even though he was not using the

85

Scrum itself as a forum for achieving his goals, as a vehicle for team awareness and for
securing face-to-face time with the software developers, it nevertheless played an
important role in Lance’s work: on our way to the Scrum, he had informed me, “I guess

the Serum was a motivating factor this morning.”

After the Scrum, Andie and another software developer, “Gene,” remained
behind to talk to Lance. The developers began by reassuring Lance that the version of
the product interface that he had been working on would indeed be changing again. At

this

Figure 7: Annotations to Lance’s reminder document

 

 

 

 

 

point, Andie wrote answers to several of Lance’s questions and drew a small mockup of
the new interface in the white space at the bottom of the first page of Lance’s reminder
document (see Figure 7) and she and Gene proceeded to explain how this interface
would function. However, Lance countered by noting that the workflow would still be

confusing and Andie responded by asking if he could give them any advice. Lance

86

gave his advice and there was some back and forth discussion of the workflow, with
considerable attention paid to Lance’s screen captures and Andie’s sketch, and many
iconic gestures as each of the three participants drew potential interfaces in the air for
each other as part of their discussion. Lance, in particular, used these iconic gestures as
he articulated his vision for a redesigned interface, in part perhaps because he was no
longer holding his reminder document or pen. After this, there was a general silence for
about fifteen seconds as the developers looked at the page and appeared to be thinking
through the implications of what Lance had said. Finally, the developers acknowledged
that there were problems with the workflow and agreed to try to correct them. Andie
drew another mockup in the large white space at the bottom of the second page, and
showed it to Lance, who approved of the change. Lance told them to keep the reminder
document to help them make these changes, but Gene noted that their answers to his
questions were on the document. So, Lance ended up taking the first page with him and
leaving the second page (the one containing Andie’s second screen drawing) with them.
On the way back to his cubicle, Lance expressed satisfaction at the way the meeting had

turned out.

Scene 2: Meeting with Wallace

Overview of the scene
This scene took place almost one month after the first scene, on the morning of
November 21, 2007. As with the previous scene, this scene occurred in three parts: 1)
Lance composing in his cubicle from shortly after 9:00 am until 10:30 am; 2) a short
Scrum meeting in the Scrum Hall; and 3) a post-scrrun meeting with a Web developer,

“Wallace,” to discuss another user interface and workflow issue. During Lance’s

87

composing and the Scrum, my positioning was similar to that in the first scene in that I
tried to position myself behind Lance or outside the semi-circle of meeting participants
gathered around the Gantt chart in the Scrum Hall (see Figure 4) in order to be as
unobtrusive as possible. In the third part, which took place in Wallace’s office rather
than in the Scrum Hall, I positioned myself in between though slightly to the side of
Lance and Wallace, so that I could see both of their faces and gestures clearly and see
the desk surface that lay between them (see Figure 8). In addition to taking field-notes
during this meeting, I also recorded video using Morae, which I later transcribed and

annotated with data about Lance and Gene’s movements, object manipulations, and

Figure 8: Wallace’s office

-
Wallace
, Researcher
, - ‘m with Laptop

/ & Webcam
Lance 3 Lance
_

Presentation

 

 

 

 

 

 

 

 

gestures. This helped me achieve a fuller picture of the action of this part of the scene

than I was able to achieve in the earlier post-Scrum meeting.

Synopsis of the Action of the Scene
When I arrived at 9:00 am, Lance explained to me the purpose of the activities
he would be engaged in. As with the first scene, he was concerned with a user interface

and workflow problem that he perceived in his product. In this case, the issue

88

concerned the manner in which this product, a Web-based product, would be integrated
with its much more complex and full-functioned boxed cousin if users happened to try
to use both on one computer. He had been made aware of the issue by one of the
customers currently involved in the beta test of the product, who had posted a message
to the beta Wiki indicating that she was confirsed by several aspects of the integrated
interface. After learning of this, Lance had grown increasingly alarmed that the
integration of the two products was not being thought through carefully enough — a
notion reinforced by a meeting of product managers he had attended the previous Friday
in which the integration was presumed to be “an uncomplicated affair.” His purpose
during this work session, therefore, was to update a PowerPoint presentation to take to
the post-Scrum meeting with Wallace, “a talented web developer,” on his team who had
particular clout as the person in charge of all the interactions with beta testers for that
product. Lance hoped, in short, to win Wallace’s support in his effort to change the
integration interface, though he acknowledged that this might be difficult because

Wallace “gets pulled in two directions a lot.”

Lance began work by opening a PowerPoint presentation that he had created on
a previous day. In our follow-up interview, he informed me that he had created and
delivered this first portion of the presentation to his information development
teammates, their manager, Becky, and their director the day before so that they would
all be “on the same page” when talking to the rest of the company about the issue. Even
though his meeting with Wallace would be one-on-one dialog instead of a presentation
delivered to a group, he had opted to update this PowerPoint instead of creating a new

document from scratch. He was planning to add slides containing sequential

89

screenshots depicting step-by-step walkthroughs of two scenarios users might encounter
when using the integrated product interface. Rather than recapturing the screenshots
himself, though, he told me that he hoped to find the screenshots, which he knew that
Monica, who had earlier been assigned to the product, had already taken and placed
somewhere on the N drive. After a few minutes of fruitless browsing on the N drive for

these, however, he gave up, muttering, “we gotta get a Wiki or something.”

He next opened his screen capture software and the product itself and proceeded
to navigate through the product’s interface, taking screenshots as he went and pasting
these into the PowerPoint. At first he was, for speed, pasting every screenshot into a
single slide, with the intention of breaking them out onto separate slides when he
finished with his capturing, but, informing me that he was afraid he would forget the
prOper sequence, he started cutting and pasting the screenshots into individual slides,
sequencing them to imitate the functioning interface for the scenario he wished to show
to Wallace. When Lance finished this, he went back and polished up his slides by
adding some explanatory text and graphics to enhance the visual appeal. Finally, he
informed me “that’s good enough for Wallace,” printed it, set the printout on his desk
where he wouldn’t forget to take it to his meeting, and began to work on another short
task in the interval before his Scrum. As before, he worked right up until his Outlook
reminder informed him his Serum was about to begin. He took his pen and printed

version of the PowerPoint and we left for the Scrum Hall.

As in the first scene, the Scrum meeting itself was quick and fairly uneventful,
with Lance using it mainly as an opportunity to remind Wallace that he wanted to meet

with him afterwards. Lance did update the Gantt chart by moving the note containing

90

the task concerned with fixing the integration issue, which he said had become “like a
cancer,” to the next Sprint so that he could cross it off his to-do list for the current

Sprint.

Immediately after this meeting, Lance, Wallace, and I proceeded to Wallace’s
office, down the hall from the Scrum Hall, in the software developers’ wing of the
office. Wallace began by asking the purpose of the meeting. Lance recapped for him
the previous Friday’s meeting in which the product integration was described as
“seamless” with, Lance believed, no consideration of the development effort involved
in making it so: “make it happen, developers, make a button or something.” At this
point, Lance placed his PowerPoint printout face up on the desk facing Wallace and, for
the next minute and half, stepped Wallace through the workflow in detail, pointing
frequently to the screenshots with his pen and alternating his attention between the

printout itself and Wallace.

After Lance finished his walkthrough, Wallace indicated that he was not
particularly worried about the issue Lance had identified, or at least that it was not a
new or unique problem to the new software they were developing. Wallace made this
point by invoking his superior knowledge of the product history: “It’s not necessarily a
new problem [. . .] to me this is not really a new thing, this is the same kind of drawback
[as in the previous versions of the product].” Wallace also made use of Lance’s
PowerPoint printout to make his points, especially when asserting that he considered
the problem more a matter of providing adequate documentation than of correcting the
workflow in the software itself. For instance, Wallace suggested “I think it’s you

choose ‘upload’ and like it gives you exactly, it gives you the synopsis” while pointing

91

to the spot on a particular screenshot where he thought such a “synopsis” or embedded

user assistance ought to go.

Lance held his ground, however, countering each of Wallace’s points by
appealing to his own knowledge of the user, and at several points even speaking in the
person of the user, as when he said, “if I’m using [the product] and I want something
and then I do all that work and I send it out there and I don’t understand if I get
surprises then I’m going to get so mad because the profile took out something or [think]
‘why is it like this?”’ Lance also attempted to convey to Wallace his own vision of
what an improved interface and workflow would look like by alternately pointing to his
screenshots and “drawing” pictures of a revised interface in the air. This attempt to
offer viable alternatives rather than simply critiquing is a strategy that Lance identified

in our initial interview as key to successfully performing his job:

One thing that I think [Software Unlimited] does well [. . .] is that it seems like
we accomplish a fair amount per minute compared to a lot of meetings I’ve been
to in my past [. . .] We usually have some sort of printed agenda [. . .] You know
— concrete, just go and just pound it out. What I’ve learned lately too is if you
don’t like something, it’s not effective just to [only critique it]. It’s good to have
alternatives and I don’t even care if they take those ideas or not, but it was good

to be able to compare apples to oranges instead of apples to nothing.

Finally, Wallace, seemed to accept Lance’s view, if a bit half-heartedly: “I can
see the benefit, now I don’t think that they are going to do this a lot, but I could see the

benefit.” The meeting concluded with Lance planning to refine the PowerPoint in order

92

to continue trying to gain traction on the issue: “So I’m going to fix this up and try to

better define that.”

In our follow-up interview, Lance seemed cautiously optimistic that he had
made his case to Wallace, and in a follow-up email about the issue that I sent to Lance a
few months later, he updated me on his ongoing struggle to make his point and he
informed me that Wallace “probably sees my point of view 75%, but is not nearly as
passionate and has other problems of higher priority for him.” But Lance also indicated
that some progress was being made, even though it was not necessarily a direct result of
his own efforts: “In related news since then, people working more in sales and
marketing that have to talk about this with real paying customers I think share my

points.”

Analysis of Lance’s Memory work

Archiving Versus Reminding

Comparing Lance’s performance of memory work in these scenes with his
descriptions of his memory work practices in the initial interviews reveals an interesting
discrepancy, a discrepancy that holds significance for understanding the ways in which
different types of memory are valued and utilized at Software Unlimited. Remember
again that, as I define them, the distinction between reminding and archiving practices
lies chiefly in that the former are principally concerned with preserving memory for a
limited term and for a single (or very limited number) of uses while the latter generally
exist to preserve memories for longer periods of time and for multiple or less clearly

demarked purposes.

93

The discrepancy is as follows. In the interviews, Lance assigns his archiving
practices a far higher degree of importance than his reminding practices while in the
scenes of memory work, it turns out to be reminding practices that consume much of his
time, attention, and effort. First, as I noted above, Lance indicated in the interview that
he considered the archival work of preserving his long term memory in his “back of the
mind” folder to be quite important because Software Unlimited encouraged and
rewarded innovative thinking and he believed that he would likely be able to use this

information productively in the future.

Perhaps even more importantly, Lance indicated that he regarded these
archiving practices as critical to his success in his current job tasks. For instance, in
response to my question asking him if he finds that the projects grow more difficult as
they proceed, he rather emphatically indicated that his archiving practices, particularly
the sequential and jussive aspects of those practices, make the job relatively easy,

freeing him up for more innovate work:

It’s not a big deal. I think like with the core deliverables [. . .] those deliverables,
those go in a very special place like it’s all perfect and that’s no problem. Then
I have this battery of stuff I might or might not use to create those and that stuff
gets kind of wild sometimes [. . .] SO like there’s this pile of stuff that seemed
insurmountable so I made up this table — this is many iterations of this — and
what it does is it sort of makes it into bite sized chunks and then I can sort of

prioritize stuff or whatever.

94

That is, Lance sees these archiving practices fueling his invention processes. At another

point in the interview, he says:

If there’s a big debate [. . .] then for that I’ll actually take the time to write up

[. . .] things that nag at me a lot I put sometimes in the ‘back of the mind.’ [. . .]
Stuff that goes in here is stuff that might someday be useful or it’s some note
that I had [. . .] like sometimes people ask me if you have anything for a session
description or for a flyer, and maybe if in the future we work with marketing
too. Anyway, this is just three sentences or two different sentences that have
pros and cons so that in one breath of air you can explain it and that’s come in

handy quite a bit

In other words, Lance uses this archive as a resource to meet the day-to-day demands of
his composing situations. He believes that this archiving practice contributes to his

success in his job.

In comparison, Lance does not assign nearly as much importance to his
reminding practices. First, whereas he talks about his archiving practices in 14
segments of our initial interview, he spends half as many talking about his reminding
practices. Further, he does not simply fail to mention his reminding practices, when he
does talk about them he seems in general somewhat dismissive of their importance to
his job performance. For instance, of the reminding tools in Outlook, such as the
Calendar function, which I was later to see him rely on in both the scenes of memory
work recounted above, he said: “I know some people use Outlook way more efficiently

than I do. I only use email. I don’t use the journal; I don’t use the other stuff.” And,

95

even though he was willing to concede that Outlook Calendar was “better than nothing”
he did not see it as very important for his own reminding purposes: “it’s really funny
how everyone is so chained to their Outlook. I guess you have to be and one good thing
is that if you have a meeting scheduled to end at two then it probably has to end at two
because they have a meeting somewhere else. I guess it keeps things from dragging on

a bit.”

Lance was similarly dismissive of another of those ubiquitous tools for
reminding in contemporary workplaces: post-it notes, telling me twice during our initial
interview that he did not use them for work very much, but, again, using them for work
purposes on several occasions when I job-shadowed him. Instead of ephemera like
post-it notes, Lance indicated that, for work-related purposes, he preferred rolling his
reminders into longer-term solutions (archiving, in other words), as the following

quotation illustrates:

Like the matter of that skip button, it consumed me and so what will happen is
I’ll just be at home or in the morning on the drive in or in the shower or
something like that and I’ll have like this list of things in my mind I don’t really
write down on post-it notes or scrap paper that much and then I come here and

I’ll write it and I’ll stick it in this [back of the mind] folder.

As I questioned Lance about this preference for archiving over reminding, his answers
indicated that memory stored as reminders lacks two characteristics that he values:
classification and contextualization. The first, classification and its attendant

phenomenon of sequencing (Bowker & Star, 1999) relate to the permanence, stability,

96

and accessibility that Lance perceives in memories stored in archives. Thus, of one of
his primary reminding practices Lance says somewhat disparagingly, “I’ve thought
about getting a little notepad you can stick up there [gesturing to the top of his laptop
screen] but pretty much my scheme for that is like if I need something short I actually
just open Word and then I usually just save it to my desktop and when I’m done I delete
it, which is probably not as efficient.” By contrast, Lance believes that an archiving
tool like Microsoft’s OneNote would allow him to store these small chunks of
information in more permanent, classifiable, and, hence, more useful ways: “supposedly
you can get a lot of stuff in one place [. . .] you know like I need a little information [. . .]
put into a little document you know? Consolidate lots of little things I guess was what I

was thinking it [OneNote] might do for me.”

Derrida (1995) offers a possible explanation for the type of affordance that
Lance believes he needs in a memory tool. Tools like OneNote would allow Lance to
“consign” his memories to safe keeping in the archive: “consignation aims to coordinate
a single corpus, in a system or a synchrony in which all the elements articulate the unity
of an ideal configuration. In an archive, there should not be any absolute dissociation
any heterogeneity or secret” (Derrida, 1995, p. 3). In a sense then, although he certainly
uses reminders, Lance prefers not to engage in reminding as a practice, preferring
instead to store his reminders in an archive where they can be standardized, classified,
and stored in an orderly structure that will render them usable if a future need should

arise.

The second aspect of archiving that Lance appears to value is contextualization.

Because reminding is, by definition, about creating transitory objects or cues, reminders

97

do not in general need to contain much information about the circumstances
surrounding their own creation, instead relying on the circumstances of their use to
provide any necessary background context. That this is a problem with reminding for
Lance became especially apparent when he discussed the memory practices of his
product team. Lance informed me that, since becoming the semi-official meeting “note
taker” for his product team, he had brought a new rigor to the job of preserving
knowledge generated in meetings so that this knowledge would remain useful. Lance
contrasted this rigor with the earlier slapdash reminding practices employed by the team
during its crucial Sprint planning meetings. Showing me an example of these earlier

efforts on the N drive, he said:

Okay, so see what someone has done here is copy the whiteboard. But see
there’s no that’s another thing too like a lot of times we can’t remember what
does ‘error this consistent with Mac’ mean? [with rising intonation] What does
that mean? Whereas with the notes, I just type like crazy and then I would like
write actually more than what would be on the [whiteboard], like the whiteboard
didn’t say that. [. . .] Next time what I’m going to do is cut and paste all the
meetings and put them on a separate page that just has meetings and then what

we like to do is get those done early in those thirty days.

Lance prefers the greater contextualization that the activity of consigning these
Whiteboards to an archive allows him. By transforming the reminders on the
whiteboard into written accounts, he ensures that what started as, essentially, lists of de-
contextualized “action items” from meetings become permanent, accessible, and

interpretable records in the company’s memory.

98

Archiving practices such as Lance’s contextualization of the whiteboard are
quite time consuming and labor intensive; they form, in short, one of Lance’s major job

activities. As Bowker (2005) points out:

The more information you provide in order to make the data useful to the widest
community and over the longest time, the more work you have to do. Yet
empirical studies have shown time and again that people will not see it as a good
use of their time to preserve information about the data beyond what is

necessary to guarantee its immediate usefulness. (p. 117)

The software developers and other members of his product team obviously did not
believe this was a good use of their time. But Lance clearly did and he was not alone in
this — his emphasis on archiving as a key element to his success was repeated by the
other information developers in their interviews, as was the de-emphasis on reminding.
For example, while discussing a relatively new practice that she had adopted of keeping
a carefully-organized notebook of meeting notes, Lucy said, “I can go back and see if
we made decisions or whatever,” which makes her more effective and influential in
meetings with her own product team: “they want me there.” Lucy similarly compared
her new archiving practice favorably with her earlier, more reminding-like method of
taking notes: “Much more than I ever used any of those notebooks that I just, you know,
had randomly. And I think it helps me be a lot more productive [. . .] I know what’s
going on and I know that I know what’s going on, so I speak up much more often than I
used to on a lot of that stuff.” Peter, too, noted the importance of archiving to his job
duties, particularly emphasizing those duties concerned with acting as an advocate for

users. Of his own complex system of utilizing multiple legal pads for keeping notes,

99

Peter says: “I find a lot of what I do is like when people have trouble designing things
or implementing things I tend to write that down because that’s going to be something a
user might have trouble with later if it was hard to design.” Becky and Monica also
expressed similar preoccupations with keeping their memories classified and

contextualized.

Only Angela, the information developer with the most seniority, differed
somewhat in her approach to memory work — and understanding this difference offers a
possible roadmap towards understanding the emphasis placed on archiving by the group
as a whole. The difference is that, where the other information developers described a
lot of careful articulation work and use of external tools to assist them in their archiving
practices, Angela indicated that she had, by virtue of her lengthy tenure with the
company, virtually transcended the need for external tools or careful articulation work:
“other people, they have file folders and they keep track of things. I just don’t.” That
is, Angela’s nearly six years with the company meant that archiving for her appeared to
have become largely internalized and that when she did employ external tools for

archiving, she did so largely to help other people rather than herself:

There are several people on the [information development] team who are very,
very structured. They like lots of lists and check lists [but] I’ve been doing this
so long and I’ve been here [. . .] a long time. I’m able to just know a lot of stuff
— historical stuff -- and why we did things. Whereas people who are new coming
on board, they don’t have that historical perspective and so they have a need for
that information and so I spend, you know, some down time if I have any

documenting things about the way they were or how we do things because it

100

used to just be ‘ask [Angela],’ ‘ask [Angela],’ ‘ask [Angela].’ So, um, I think I
try to remember as much as possible and we have a place where other people

can go for stuff, if they can find it.

Angela’s attitude toward archiving applied equally to her reminding practices. For
example, she informed me that she does not take notes at meetings, preferring instead to
doodle on a piece of paper: “Doodling frees up some place in my brain that helps me
free think on something that’s being discussed. In some way doodling helps me focus
more on the conversation and think through the complicated processes.” Similarly, of
everyday reminding practices like creating to-do lists, Angela noted her lists live only

“in the back of my head.”

These were not boasts: Angela’s grasp of the “historical stuff” and the
effectiveness of her memory were confirmed by every one of the other information
developers. Lucy, for example, informed me that Angela had been her mentor and
expressed awe at the extent of Angela’s knowledge: “And I think that one thing that
I’ve noticed here is all of the reasons — back reasons — Angela’s been huge — I can’t
even imagine going into that mind for all the stuff she has in there. Angela could
explain to me the three year reason — you know the reasoning over the last three years
of how something had gotten...” Similarly, the information development manager
Becky asserted, “We are really lucky because she’s got so much background and
understands sort of how we got to where we were with some of these choices and
different things that were tried. So we are really lucky to have that. Some of the teams

don’t have as much history.”

101

F entress and Wickham (1992) help shed light on Angela’s internalization of
archiving practices in the following passage in which they articulate a concept rather

similar to the standing wave of memory I introduced in Chapter 2:

If, in certain situations, memory seems fiagile and volatile, it is because so much
of our memory is the memory of context-dependent information. As long as we
remain in these contexts, we remain surrounded by clues which prompt our
memory. Here, in what we earlier called our ‘present continuum’, the external
environment itself takes over the job of ordering memory into a sequence: we
remember things in the order in which they habitually appear. [. . .] Take away
the external, material support, and the memory of the ordered sequence tends to

fade. (p. 73)

An important source of this context for Angela lies in her involvement with the same
product for six years: “One of the goals that has been in place since I started with
[Software Unlimited] was design a [. . .] product that Angela can actually use. [. . .] I am
the first customer.” In short, working in the same physical and informational context
for so many years has helped Angela to develop a hexis, which has largely supplanted
the need for the archiving tools used by her newer colleagues. Carruthers (1990) has

this to say about such a hexis:

Quintilian defines hexis as that ‘assured facility’ (firmafacilitas) in any art
which supplements and transcends the rules themselves, and constitutes what we
call mastery. [. . .] Hexis is physiological, as the memory is trained to respond to

certain movements, just as a dancer’s muscles are, but it is also reasoned, for it

102

is a ‘facilitated’ rather than ‘automatic’ response. [. . .] For the trained memory
was not considered to be merely practical ‘know-how,’ a usefirl gimmick that
one might indulge in or not [. . .] It was co-extensive with wisdom and

knowledge. (p. 69)

Finally, Angela’s hexis contributed not only to her memory practices but also to her
authority in the company, a consideration which takes on additional significance when
read in light of the discrepancies I noticed between my participants’ accounts of their

memory work and the ways in which that work was actually practiced.

Reminding Versus Archiving
So, to reiterate the discrepancy that I noted between participants accounts of
their work practices and those practices themselves: I noticed that despite all their
dismissing and downplaying of their reminding practices compared to their archiving
practices in our interviews, in the observation sessions in which I observed my
participants actually engaged in work, reminding appeared to play a far greater role than

my participants had given it credit for.

This was particularly true of the two scenes of Lance recounted in this chapter.
In the first scene, recall that while he was working on the user guide for his product,
Lance created a Word document in which to write the list of questions he wanted to ask
the software developer, Andie, after the daily Scrum. Before he ran off to the Scrum, he
printed this document and saved it to his desktop, and then, at the conclusion of the
meeting with the developers, he offered to leave it with them. These two actions —

saving the file to his desktop (the place where he stores temporary reminders - “they

103

[. . .] find their way to the recycle bin pretty quickly”) and his willingness to part with
the hardcopy after it had fulfilled its purpose - indicate to me that the creation of this
document represents a nearly ideal example of a reminding practice: the document was
ephemeral, it was created in order to serve a single purpose, and it did not have a future
in any of the company’s official archives. By comparison, the user guide Lance was
writing during the same scene would seem to have been by far the more important
document — it was an official text, a “deliverable” of the type that Lance made sure to
store in “a very special place” on the N drive where “it’s all perfect.” Yet, in
comparison to the reminder document, the portion of the user guide that Lance was
working on during this scene was not only not as visually complex, containing no
screenshots, but it also would, if the reminder document did its job, be radically

changed or replaced when the software interface changed.

Thus, the archived document ended up being the more ephemeral and less
important of the two documents: the reminder document would live on in the interface
while the user guide would be supplanted. Further, the reminder possessed affordances
for responding to a particular scene of kairos that the user guide did not. It travelled
where a user guide could not easily travel and it spoke to the developers in their own
language. In this meeting, therefore, the reminder document served as an object of joint
cognition, an “(im)mutable mobile” enabling Lance, Andie, and Gene to invent together
(Latour, 1987). The document would remain invisible and would not live on in either
Lance or the company’s memory, but the product of its labor would. In sum, reminding
practices in that scene appeared to play a far greater role in Lance’s work activities than

he realized.

104

While it seems certain that the undue emphasis that my participants placed on
archiving over reminding activities in their interviews was at least partially a result both
of my interview questions (see Appendix) and of the manner in which I had
characterized the purpose of my research to them, this undue emphasis was also
possibly an indication of what my participants believe counts or qualifies as memory
work in their organization. That is, I would argue that it reveals what Lance and my
other participants drink of when they are asked about to describe memory work: stable
structures and databases rather than the (im)mutable mobiles and ephemera of
articulation work. Further, this behavior might be an indication of the memory regime
in place in this organization, a regime that privileges particular forms of knowledge

over others. The second scene raises several issues that help illustrate this point.

Although the PowerPoint that Lance took with him to his meeting with Wallace
in the second scene was intended to serve multiple purposes and to live on in the
company’s archives, and hence was not purely a reminder, reminding was
unquestionably the main purpose of Lance’s work on it during this scene. Certainly, it
appeared that his addition of the screenshots was intended primarily to serve as a
reminder to both himself and to Wallace as they tried to hash out the issues: “that’s
good enough for Wallace,” he said when he finished updating it just prior to the Scrum.
That is, even though the PowerPoint was intended to communicate and persuade
Wallace, it was also, though perhaps a bit less obviously than the document in Scene I,
intended to serve as reminder document so that Lance could remember the points of his

argument and Wallace could follow along. I would, therefore, characterize the memory

105

work that Lance performed in this scene as primarily an example of reminding rather

than archiving.

As Lance and Wallace worried over this PowerPoint, passing it back and forth
several times during the meeting, a key theme that emerged was that Lance and Wallace
considered different types of knowledge as warrants for their arguments: Lance
appealed primarily to knowledge of the user while Wallace appealed primarily to
knowledge of the product itself. That is, in both his initial resistance to and in his
eventual (grudging) acceptance of Lance’s critique of the product interface and
workflow, Wallace primarily invokes his superior knowledge of the product and its
history. He, for example, suspected that the issue Lance was raising was not a new
problem, “to me this is not really a new thing” and, if it was, he believed that it could be
solved by simply finding some real-estate on one of the user-interface screens in which

to insert embedded documentation.

Wallace was, it appears, invoking Lucy’s “back reasons” or Angela’s “historical
stuff” to warrant his arguments. This “historical stuff” is important, and it lives quite
concretely, though often unnoticed in Software Unlimited’s archives and
infrastructures, particularly, as this scene shows, in its product interfaces. Bowker
(2005) makes an observation about programming languages that applies equally well to

Software Unlimited’s products:

A programming language that operates as part of an organizational infrastructure
can have potentially large effects on the nature of the organization [. . .] So

object orientation is on the one hand a model of the world, and on the other hand

106

the world is learning how to model itself according to object orientation. This
kind of bootstrapping process is common when you deal with infrastructures.
[. . .] The programming language is very much part of the organizational history

and vice versa. (p. 31)

In other words, as both the primary objects of attention of the company’s employees
and as the products of their ongoing labor, Software Unlimited’s software products
themselves hold perhaps the most important memories of the organization. That is, the
“back reasons” exist in the present in the evolving products themselves, a circumstance
which both preserves and obscures these memories: “you don’t need a memory if

evolution is doing your thinking for you” (Bowker, 2005, p. 100).

The back reasons, held in the company’s products, determine which warrants are
justified and which are not justified. Newer employees or comparatively lower-status
(than the software developers) employees like information developers, such as Lance,
then, believe that they must constantly strive to consign these memories held in
products and interfaces to usable archives by classifying and contextualizing them.
Hence, it is not too surprising that, when they are asked to talk about their memory
practices, information developers think in terms of stabilizing and permanently
preserving memories —- that is what the memory regime implicitly requires them to do.
Or, as Bowker (2005) again puts it: “It is not just the bits and bytes that get hustled into
standard form in order for the technical infrastructure to work. People’s discursive and
work practices get hustled into standard form as well. Working infrastructures
standardize both people and machines” (p. 112). At Software Unlimited, this “standard

form” is principally determined by the tools and infrastructures of the software

107

developers — as Lance points out, “it’s a very developer-centric work environment,” so

what counts as memory for developers counts for the other employees as well.

In contrast to the comparatively solid and concrete knowledge of products
exhibited by interfaces or held in more senior (like Angela) or higher status (like
Wallace) employees’ heads, knowledge of users of the type that Lance uses as warrants
for his arguments is “homeless” and without concrete representation in the organization.
To illustrate this: when I first began this study, I was quickly struck by the emphasis
that Software Unlimited seemed to place on users and the user experience. Usability
testing and product demonstrations for customers were, apparently, regular occurrences.
This led me to ask Becky if drere was any single repository or archive of knowledge
about users in the company and, if so, who “owned” this information. She answered in
the negative: “I drink you just have lots of little sources for it when you are doing your
work. It would be amazing to have that sort of one stop place to get all that information
for research or whatever but I just don’t know that anyone has that as their job. It’s sort
of all of our jobs but we do it for different reasons and for different things.” That is,
unlike product knowledge which is literally placed before the eyes of virtually every
employee every day, knowledge about users exists in a much more ephemeral state in

this organization.

This phenomenon of dispersed knowledge about users was reinforced by the
other information developers. For instance, when I asked Monica if user-testing data
was made available to her, she said that she and the other information developers were
free to watch the user testing as it occurred, but she seemed to imply that more

permanent records or detailed studies were not generally made available to everyone but

108

instead had to be acquired from the group conducting the test: “They show us an
overview and if we wanted to see more details or if I had specific questions about
something I could go and ask someone to show me their findings and research or data.”
In fact, the issue of which groups were permitted to conduct user testing became a
source of controversy during the period of my study. The information developers
conducted their own user test of a product but doing so led to a conflict with another
group over who possessed the right to conduct user tests of products. The information
developers eventually ceded the issue and did not conduct any further user testing

during the study time-period.

So, lacking full access to the back reasons concerning either the products or the
users, newer information developers like Lance must perform a great deal of
articulation work to achieve their goals. Often this articulation work takes the form of
efforts to stabilize, contextualize, or embody knowledge about users in some fashion.

In Lance’s case, he sometimes performed this work through gestures, as he traced his
vision of the interface in the air, he sometimes created reminders for himself or his
interlocutors or to serve as objects of joint attention, and he sometimes even acted in the
persona of the user. In each instance, however, this articulating entailed a great deal of
memory work to help Lance share his phantasms with other people so that he could

convey his vision of the user experience.

Conclusion

These scenes of Lance’s memory work have begun to delineate some of the
details of the memory regime at Software Unlimited. Understandably, this memory

regime seems to most highly value knowledge of its software products. However, this

109

knowledge, though it resides in the product interfaces and is drerefore in a sense
universally available to everyone in the company, is nevertheless inaccessible to many
employees because the constantly-evolving product interfaces do not describe the
circumstances and contexts of their own creation in a language that is equally accessible
to everyone. The information developers, particularly newer ones like Lance, then,
seem to expend a great deal of effort on articulation work to contextualize this

knowledge and place it into a system where it can be readily found if needed.

In addition to product knowledge, these scenes also revealed user knowledge to
be important to this organization, especially to the information developers, for whom
user advocacy is a principal job duty. It was in trying to transform this domain of
knowledge so that it would be visible and comprehensible to other employees that
Lance appeared to rely most on his embodied contexts for invention. That is, it
appeared that it was to solidify and give a “place” to user knowledge that compelled
Lance to employ all the tools at his disposal to bridge the gap between his position and
that of his interlocutors. In other words, Lance was attempting to raise the status of user
knowledge as a topoi to the same level as that of product knowledge as a basis for
arguments. The limited measure of success he experienced in these efforts is some

indication of the obduracy of memory regimes.

110

CHAPTER 5: TWO SCENES OF MEMORY WORK WITH LUCY

Introduction

In this chapter, I focus on several breakdowns and recoveries drat Lucy faces in
two scenes of memory work that I observed. I will argue that these breakdowns and
Lucy’s efforts to recover from them further delineate the memory regime at place at
Software Unlimited, a regime that exerts specific demands requiring specific types of
“street smarts” or articulation work of its subjects (Star, 1999, p. 387). In Chapter 4, I
focused on what might be thought of as the “input” aspects of this memory regime: its
influence on the archiving and reminding practices by which information developers
attempt to transform and stabilize certain types of knowledge for later use; in this
chapter I will focus more on the converse, the “output” aspects of the memory regime,
in which it influences the behaviors of the information developers as they attempt to

retrieve stored memories via the activities of finding, referencing, and gesturing.

At the time of our interview Lucy had been employed at Software Unlimited for
about two years, giving her a fair degree of seniority in the company. However, illness
and the birth of her second child had kept her on leave or working part time up until one
month before our initial interview when she had returned full time and begun serving as
the “lead writer” on a product team once again. Before joining Software Unlimited,
Lucy had earned a university degree in technical communication and had worked for
five years at a much larger company as part of a team of around 100 technical
communicators. This history meant that Lucy has had to make a number of major

adjustments to her work patterns in recent years: adjusting to working at a much smaller

11]

company, adjusting to working as the only information developer on a product, and

adjusting to balancing full-time employment with the task of parenting two children.

These adjustments played a major role in Lucy’s memory practices, and she
appeared to be the information developer with the most carefully-planned system of
organizing and preserving her memories. First, her comparative newness as a full-time
employee and lack of experience at a company as small as Software Unlimited, strongly
motivated Lucy to try to keep “historical stuff” readily at hand. That is, similar to
Lance, Lucy engaged in several carefirlly thought-out archiving practices. One of these
practices, which she alluded to as a “data dump,” entailed keeping carefirl records of
each stage in the evolution of her product, including critical contextual information

about what precipitated each evolutionary step in the product’s design:

One thing that I’ve noticed here is all of the reasons [. . .] the reasoning over the
last three years of how something had gotten . . . There was a lot of stuff like
that around here — you know just being a small company with all this stuff it was
like all drese reasonings, so I think that’s why I even started collecting a lot of
drat, the reasoning part, because it makes a difference around here a little bit. I
don’t know if we are kind of getting away from that as we get a bit bigger.

People expect that you know the reasoning, like why it is the way it is.

Lucy organizes this data dump archive by Sprint and by product feature in order to keep
track of every stage of the product’s evolution: “1 have a ‘Sprint’ folder and dren in
drere I have different things but I have one folder called ‘Features’ and in the folder

features I have a different folder for each of the things that we are working on.”

112

Second, as anodrer consequence of the size of Software Unlimited in
comparison with her earlier job, Lucy has had to learn to deal effectively with the
relative lack of structure and increased independence that her current situation gives
her. Where archiving tasks were performed automatically in the previous job, at
Software Unlimited Lucy has to be thoughtful and proactive about them: “we had very
specific things that we did. We had to keep drings. Our computers were backed up
every night. We didn’t have to do anything. I didn’t have to do anything then.”
Further, because there were almost 100 technical communicators in her earlier job, there
was a degree of management oversight that left them with little need - or ability — to

innovate or to manage information themselves:

It was very disconcerting for me when I started working here that I didn’t have
templates. That nobody had templates they were giving me. I considered that I
could work pretty independently — and I could — but it was really much more
than I thought it would be [. . .] having learned that I wanted those things and I
remember complaining about them when I had them. You know, having to fit in
to those things, but it’s really kind of comforting. So I think the more stuff like

that I have to do the better I am at tracking drings.

Despite the initial discomfort she expresses in this quotation, Lucy had gained
confidence in her system for managing information. She had, in short, become “really
proactive” because “no one else is going to track it.” In fact, Lucy’s organizational
practices had come to be admired by the other information developers as well. Peter,
for instance, noted: “She’s really great at being organized. She works very differently

from me. When I worked with her it was enlightening to see how she worked and then

113

I tried it and it didn’t quite work but it was really good to try that. She’s very organized

and has detailed documentation plans and likes to do tables and is very organized.”

One of Lucy’s major innovations in project tracking was the “documentation
library plan,” a new genre at Software Unlimited that Lucy herself had introduced to the
company and, consequently had been given the privilege of naming. This document

served a variety of audiences and purposes, bodr memory and communication related:

So we had this huge [. . .] we put everything in there [. . .] and then I go through
and each feature I have a table that says the name of the feature [in] the heading
[. . .] That’s kind of the first part —— describing the feature with enough
information so that the developer could know that we know what this feature is
going to be [. . .] The whole reasoning behind it was so that I could go drrough it
later and I could see that my first initial instincts were for how this feature was

going to affect my help basically.

Lucy attributed the idea for this document to her experience at her previous employer,
and Becky, the information development manager, indicated that, although she allows
each information developer considerable latitude in its contents and format, the
document had achieved the status of a genre at Software Unlimited and become a
standard part of the team’s process for beginning a project “so if the next writer picked
it up, they’d have a history. All of drem do it a bit differently but the idea is to sort of
present to the team ‘here’s what I think we are going to have to do to support this

product.”’ In sum, Lucy’s memory practices were considered so effective that they had

114

influenced those of the team as a whole; Lucy was regarded as the organizer par

excellence among the information developers.

Finally, her busy home schedule (which meant that she worked from 6:30 am
until 3:00 pm each day in order to pick up her children from school) combined with the
frequent meetings required by her job as the lead information developer on her product
team compelled Lucy to dedicate considerable attention to her reminding practices
every day. Significantly, she couched the need for these reminding practices and the
difficulty in finding time to get her work done by invoking an embodied metaphor from

the domestic realm:

I do better when I can concentrate on drings for a long time rather than doing a
little bit here and there. And that’s really what my schedule helps me with is
that I know — once I’m started on something I’m good — but lots of times if I’m
not getting started on somedring I kind of can’t concentrate. It’s like you go and
find a cup in the living room and you take it in the kitchen but on the way you
see something in the dining room and then you start cleaning that — you know,

that kind of thing.

Lucy keeps this schedule using Microsoft Outlook Calendar, which she uses in a far
more structured way than any of the other information developers. She essentially
schedules almost every hour of every day in the Outlook Calendar, including her blocks
of “free” time, the descriptions of which she labels “private” so that people will not
attempt to arrange additional meetings during the precious blocks of time that she uses

for extended composing tasks. At the end of the week, she also uses her Outlook

115

Calendar to remind both herself and others of her progress during the week: “it helps to
jog my memory for what I really did and to be able to show it.” The Outlook Calendar,
then, serves multiple functions for Lucy: it is, of course, a memory tool with both
reminding and archiving affordances, but Lucy also employed it as a communication

tool to demonstrate to her colleagues drat she has worked as hard as they have.

Lastly, two other points are worth keeping in mind as I discuss the scenes of
Lucy’s memory work. First, unlike Lance’s product, the product to which Lucy was
assigned as the primary information developer was an older “boxed” product. And
second, like all of the other information developers, Lucy identified her role on her
product team more as one of user advocating than of writing: “I think the biggest roll
that we do is user advocate. I really don’t drink that writing per se — I think that’s the

smallest part of our job.”

The scenes recounted below were recorded in January and February 2008, using
the webcam and software recorder features of Morae. Additionally, I was able to
simultaneously view the proceedings from the office assigned for my use (see Figure 2)
using Morae Observer, which streamed the video live over the company network. As I
watched, I took notes about Lucy’s actions and noted anydring that I did not understand,
and, using these notes, I conducted short follow-up interviews with her. Not
surprisingly, using Morae helped me to achieve a much more finely-grained account of
Lucy’s work activities than I was able to achieve through joumaling alone in my
observation sessions with Lance. Further, because she frequently worked with a

webcam pointing at her and was frequently recorded with Morae as part of her normal

116

job duties, it also likely allowed Lucy to work in a more unselfconscious and natural

fashion than if I had observed her directly.

However, I did not find the experience of conducting research remotely to be
entirely satisfactory. For instance, because of the lack of physical proximity to Lucy, I
was unable to ask for clarification as she worked, nor, understandably, did she volunteer
this information to the camera, as Lance had done to me in person. SO, I relied on my
Morae Observer annotations and handwritten notes to a far greater degree in
formulating follow-up questions for Lucy than I had with Lance. This made it more
difficult to interpret some of her actions later, and, while undeniably adding to my
ability to analyze her moment-by-moment actions as she composed, it also contributed
to some loss of contextual or motivational detail. In short, I believe conducting
observations in this way deprived me somewhat of an emic or insider perspective such
as I achieved when observing Lance due in large part to his many explanatory
comments and asides in which he bounced ideas off of me. In other words, I felt myself
to be much less of a coworker and much more of a researcher with Lucy than I did with

Lance.

As with the scenes with Lance, I selected these days and times for work session
observations and recordings for the following three reasons: 1) Lucy indicated to me
that these would be convenient times for me to observe; 2) her consideration of the
sessions’ relevancy to my project (i.e., she planned to engage in extended composing

tasks); and 3) my own schedule.

117

Scene 1: Updating the User Guide

Overview of the Scene

This scene took place the morning of January 17, 2008. When I arrived at
around 9:00 am, I found that Lucy had placed a laptop running Morae Observer in the
cubicle set aside for my use by the company, which would enable me to watch remotely
as she worked. Although I knew beforehand drat Lucy had Morae installed on her
computer, I had originally planned to observe directly from her cubicle as I had with
Lance so that my observation methods would match. However, even though she
offered me the choice of watching remotely or directly, I opted to use Morae Observer

since she had clearly gone to some effort to set it up.

Lucy’s actions during this scene occurred entirely at her desk in her cubicle,

facing her computer monitor (see Figure 9). Her goal, she informed me during the

Figure 9: Lucy’s office

 

Computer &
,/ Webcam

H

 

 

 

 

 

 

 

 

 

 

Lucy Q

follow-up interview, was to continue her work on a user guide that she was creating for

 

 

 

the next release of her product. The purpose of this particular user guide was to help
people who were new to the product, so Lucy wanted to make the guide as short and

streamlined as possible. As source material, Lucy was using the large Word document

118

that served as the source file for the online help for the current version of her product.
She was copying and pasting text from this online help document into the new user
guide, shortening or simplifying it to make it more appropriate for beginners, updating
the terminology and procedures where necessary to match the new release of the
product, and adding helpful notes giving readers suggestions for online help topics that

they could consult if they needed additional details.

Synopsis of the action of the scene

Lucy began work by browsing in Windows Explorer to find the current Sprint
folder under My Documents. She was looking for her working file, a chapter from the
new beginner user guide. She navigated to the Sprint folder quickly and then ran her
mouse cursor down the list of files as if running her finger down a list on a piece of
paper. Apparently she was unsure which chapter she needed to work on that day,
because she soon switched to her Outlook Calendar, found the calendar entry for the
current block of time, which was labeled “Work on Method.doc Stewart Observing,”
and viewed the details for the time block, which contained the text “Start at
C:\\Work\Prerelease\[product]\Sprint 30006\Methods\Sections\03_Planning.doc.” She
copied and pasted this path into the address line of the Web toolbar in Word in order to

open the file.

After the document opened, Lucy scrolled drrough the document until she came
to a block of text highlighted in yellow. In the follow-up interview, Lucy explained to
me that she used Word’ 5 color highlighting such as this one to remind herself where she
needed to start working, adding that she preferred the highlighting feature over Word’s

Comment feature because highlighting enabled her to select additional colors for

119

different reminding tasks. She removed the yellow highlight and then began to read the
text for the next several seconds. As Lucy read, she engaged in the gesturing behavior
that I introduced in Chapter 3 as “touching.” Recall that I defined touching as moving
and hovering the mouse cursor over a portion of one’s computer screen while reading
text or viewing information where manipulating the software does not appear to be the

goaL

After working on this document for about two minutes, Lucy apparently realized
that she needed the Word document containing the text of the online help because she
left a gap in her text-in-progress and began to look for this second document. She relied
on browsing for this finding activity by opening Windows Explorer and navigating up
and down the Sprint file structure until she found the file she was looking for, a process
which took about 30 seconds. As she browsed, she leaned in towards her monitor and
rested her chin on her left hand, a movement that she was to repeat at least four times
during my observations. Three of these four instances occurred when she appeared to
be having trouble finding a file via browsing; the other instance occurred when she was
intently reading a large block of text. I interpreted these movements, then, as signaling
that Lucy was dedicating extra attention to onscreen information, and paid special

attention to them in my analysis as moments in which she was expending extra effort.

After opening the online help text compilation document, a very large file of
several hundred pages, Lucy performed a logical search for a particular word using the
Find and Replace function of Word. This took her to a specific section of the document.
In this section, she found the text she wanted and copied-and-pasted it into her first

document. After copying the text in, she revised the wording to reflect the new context

120

and the new version of the product. She repeated this process several times over the

next 15 minutes.

At this point, Lucy apparently realized that she needed a third document because
she abruptly stopped typing in the middle of a sentence and began trying to find a file in
Windows Explorer. It seemed that this process took more effort than she had
anticipated because it took an extended time to find the right folder in the folder tree
(about 45 seconds), during which she again placed her chin on her left hand. An
additional indicator of effort was drat she apparently first Opened the wrong file because
she quickly closed it and opened another. This newest document was only three pages
long, and she found the section she wanted by scrolling rather than by using Find and
Replace. She intently read this section for several seconds, again touching the text as
she read before returning to her text in progress. She then proceeded to take out the
gerund (—ing) constructions from the section she was writing, signaling that she was
using the drird text as a style guide rather than as a source of content. Though she left it

open, this was the only time she used this drird document.

Lucy continued working as before, switching between her first document and
the help compilation document, until an Outlook Calendar pop-up reminder appeared at
9:55 am, reminding her that her daily Scrturr would begin in five minutes. Lucy then
performed several steps to prepare herself for her next work session. She first
highlighted the section she was working on in her document in progress so that she
would know where to begin work next time. Next, she created a comment for herself
on another section, writing “does this need an intro section?” Finally, she saved the file

and left to attend her meeting.

121

Scene 2: Attempting to Solve a Problem

Overview of the scene
This scene took place the afternoon of February 29, 2008. With one exception,
noted below, Lucy’s actions during this scene occurred at her desk in her cubicle. Her
goal, she informed me during the follow-up interview, was to finish creating the Word
document drat was the source file for the online help of a new feature in the upcoming

release of her product.

Synopsis of the action of the scene

Lucy began by attempting to find the Word document that she planned to work
on — the document that served as the source file for the online help of a new “Export to
Word” feature of her product. Similar to the first scene, she started by browsing
through the contents and subfolders of her “Prerelease” folder on her local drive using
Windows Explorer. She did this for about a minute without success before performing
a logical search using the Search function of Windows Explorer. The logical search
also failed to return any results, so she resumed browsing. After about 4 minutes spent
in unsuccessfirl searching and browsing, she selected a help file source document for a
similar feature, “Export to PowerPoint,” opened it, replaced all references to
PowerPoint with Word using the Find and Replace function, and then selected “Save

as,” renaming the document for the new feature.

She then opened Quadralay Webworks, the standard software used by the
Information Development team to generate its help files. She created a new project,
added the “Export to Word” document she had just created to be the source file for the

online help project, and prepared to save the project file. In the save dialog box, she

122

first browsed to the folder of the most recent Sprint, “Sprint 3006.” After pausing for a
second, however, she navigated one level up to the folder “Prerelease” and created a
subfolder for a new Sprint, “Sprint 3007,” into which she saved the project. While she
waited for her help file to generate, Lucy spent several minutes updating her Outlook
Calendar. Interestingly — and in keeping with the role Lucy said the Outlook Calendar
plays in helping her track her activities —- she retrospectively updated one event that had
already occurred by expanding the planned duration in order to, presumably, more

accurately reflect the actual duration for reporting purposes.

When it finished generating, Lucy opened the online help file. It appeared that
she instantly recognized that something was wrong with the file because she
immediately began browsing through the menus of the Webworks application. She was
apparently unable to locate the information she was looking for in the menus, so she
opened the Webworks online help. As soon as the help opened, she performed a logical
search for a particular term. This search returned about 20 topics, ranked in descending
order of match. She quickly selected the first topic in the list and proceeded to read it
for almost an entire minute. The topic was text-heavy and contained screenshots and
Lucy read it with some care, touching the text at various points as she read and scrolling
down and then back up several times. Apparently dissatisfied, she opened two more
topics, which she quickly scanned. Altogether to this point, she had spent almost five

minutes trying to solve her problem.

She next opened the Web Works website where, after scanning the home page,
she “hovered” her mouse cursor over the entries in the left navigation bar to deploy

pop-up submenu options, eventually selecting the FAQ page from one of these pop-up

123

menus. She scanned this FAQ page for about one minute before switching back to the
Web Works application, where she continued to explore her project settings. Next, she
opened the Web Works Wiki where Web Works users are able to contribute to the general
store of knowledge about that application, but again she was unable to locate the
information she needed, through she both browsed and searched the site. She continued
switching between these four sources — the two websites, the online help, and the

application settings — for the next ten minutes.

At this point, Lucy left her cubicle and went to consult Monica in her office to
see if she could help solve the problem. Monica informed Lucy that she was having a
similar problem finding some critical information herself and suggested that they meet
on a future day to discuss Lucy’s problem. Lucy returned to her office where she spent
the next 40 minutes alternately searching the websites, adjusting her project settings,
tinkering with the Word docrunent, and regenerating the help file. At one point during
this period, Lucy made extensive use of a particular Wiki page, which appeared very
technical and contained several step-by-step procedures. Lucy read this page very
carefully with a great deal of touching before using one of the procedures on the page to
modify her Word document. While following this procedure, she carefully referenced

the Wiki page by alt-tabbing between the browser and Word at each step.

Still apparently unable to resolve her problem, Lucy eventually bookmarked the
Wiki page, placing it in her browser bookmark folder “Doc Tools” and renaming it so
that it was descriptive of the problem she was attempting to solve. She saved the Word
document to the N drive, appending “_draft,” to the file name and sent an email and

hypertext link to this file to a software developer, asking him to review its contents for

124

accuracy. Finally, she concluded this activity by shutting down Webworks, taking one

last look at the Wiki page, and closing her browser.

Analysis of Lucy’s Memory Work

Breakdowns and Recoveries

Aldrough each of her colleagues praised the special care with which Lucy
organizes her working life and keeps her memories in order, the two scenes recounted
here are perhaps most interesting for the breakdowns in Lucy’s memory practices that
occurred during them. I would argue, however, that these breakdowns do not imply that
Lucy’s colleagues had underestimated her organizational abilities or that they were
symptomatic of carelessness or disorganization on Lucy’s part. In fact, Lucy actually
demonstrated the careful information management practices that she and her colleagues
had alluded to in their interviews: she took care in naming files, she maintained a
meticulously well-planned and coherent folder structure, and she made excellent use of
her Outlook Calendar for reminding and archiving tasks — all of which serves to make
the breakdowns that occurred in spite of such careful preparations all the more

interesting and insightful.

To recap the breakdowns in the two scenes, recall that in the first scene Lucy
experienced difficulties finding each of the three documents she worked on. In all three
instances she was able to overcome these breakdowns relatively quickly by using the
skills she has developed while working at Software Unlimited. For example, in the first
breakdown of this scene, in which Lucy was unable to find her document via browsing
in Windows Explorer, she used her Outlook Calendar entry, into which she had copied

the file path, to quickly recover from the breakdown — a clever redundancy in her

125

memory practices that, I would argue, is a definite indicator of the kind of “street
smarts” that Star identifies. Recall that Star (1999) maintains that “small disruptions in
the articulating processes may ramify throughout the workflow of the user” but that
expert users have, by virtue of their experience, learned to recover from these
disruptions with a minimum of difficulty (p. 387). It appears, then, that at some point in
her previous experience, Lucy had encountered a similar situation in which she was
unable to quickly locate a file via browsing and had, therefore, begun employing her
Outlook Calendar as a method of keeping the precise location of any information she

knew she would need ready at hand during a given span of time.

Similarly, although Lucy was ultimately unable to locate the document she
wanted to work on at the beginning of the second scene, she was able to recover from
the breakdown by using Find and Replace to modify a similar document to serve in its
stead. In the follow-up interview, she informed me that this was not a major problem
because the effort to Find and Replace the feature title was not measurably more
difficult than it would have been to update the existing document to reflect the latest
version of the product: “And see what took me a while was I was looking for one [the
document “Export to Word”] — I know I did one for Word, I just couldn’t find where I
did it and it wouldn’t be up to date anyway, so I left it.” In fact, Lucy indicated in the
follow-up interview that, similar to Lance, she prefers to over-document or over-
populate her archive rather than to leave things to chance, even though this can

sometimes make finding the right file more difficult:

Each Sprint we were working on stuff. So I just started ‘boom’ here’s what was

going on with this [feature] at that time with a screenshot. [. . .] I have a tendency

126

to — like I absorb drings and then I just write. I’ve always been good at just

writing like dumping the stuff out of my head onto a piece of paper.

In our initial interview, Lucy had characterized this process of “dumping stuff out of my
hea ” as an archiving practice and indicated that it helped her both internalize and
extemalize product knowledge. That is, Lucy believed that repeatedly writing about the
product both helped her learn the product and gave her a usable (i.e., searchable and

browsable) external resource for invention:

So and then I’m documenting drings as I go along. They are not the final
implementations of how this is going to work, but what I try to do is capture
what has gone on during the month we are working on something so that the
next time we work on it I know where the product is [. . .] It’s kind of a data
dump of everything I know about this feature that we are doing right now with

screenshots of where it is so that the next time I get to it I know right where...

Lucy appeared to view this data dump process as a necessary step in developing in
herself the “assured facility” or hexis that Angela possesses (Carruthers, 1990, p. 69).
In other words, similar to Lance’s “back of the mind”, Lucy’s “data dump” was her
method of learning and recalling the “back reasons” and “historical stuff” that would
make her a more effective employee in this organization. However, as the occurrence
of so many breakdowns in the two scenes illustrates, the memory regime that compels
Lucy to preserve so much detail exacts a price in time and effort. Where the scenes
from Lance’s working life help illustrate the time and effort required to store this

information, Lucy’s help illustrate the time and effort required to retrieve it.

127

Ejfects of the Memory Regime on Retrieval: Finding, Referencing, and Gesturing
So, what did Lucy’s work processes reveal about the articulation work required

to retrieve information in this memory regime? First, while two Observation sessions of
a single participant obviously do not allow me to propound any firm rules that would
apply to the organization as a whole, I nevertheless observed drat Lucy’s retrieval
preferences appeared to follow a pattern and that this pattern (or a quite similar one)
was discemable in the interviews and job shadowing data I collected from my other
participants. This pattern can be summarized in the following statement: when
performing routine work tasks, Lucy prefers location-based browsing over logical
searching and over referencing other people (i.e., asking them for help) except when she
perceives that time is a major factor, such as when a progress-stopping breakdown
occurs or when the sheer volume of information appears to be overwhelming. As I will
show, this order of preference is, at least partially, shaped in response to the exigencies

of the memory regime of this organization.

Consider the seeming inefficiency of Lucy’s preference for location-based
browsing over logical searching or referencing. Browsing can be, as Barreau and Nardi
(1995) note and as Lucy’s breakdowns illustrate, less precise and more time-consuming
than logical searching. Yet, Lucy began every instance of routine information retrieval
in the two scenes by location-based browsing: in order to locate all drree documents in
the first scene and her working document in the second scene she began by navigating
through her folder structure in Windows Explorer. Even when attempting to overcome
breakdowns, Lucy showed a preference for browsing over searching, such as when she

consulted the Web Works website or when she browsed the Web Works application

128

interface. While this behavior was doubtless partially influenced by the affordances
Lucy perceived in Windows Explorer and in the Web Works website, both of which
emphasize browsing over searching, it also seems contrary to the best interests of
someone who places a high value on squeezing the maximtun productivity out of every
uninterrupted block of time in the workday, as Lucy’s initial interview revealed. In
fact, Lucy did not hesitate to use logical searching in moments in which she seemed
overtly conscious of the need to save time, such as when she needed to locate specific
text in a lengthy Word docrunent or when she wanted to find an answer to her problem

quickly in the Web Works online help.

Why then, given the number of search tools available to her and the comparative
speed and accuracy of such tools, does Lucy appear to prefer the slower and less precise
method of location-based browsing for her routine finding activities? To answer this, it

is worth revisiting Star’s (1999) explanation of articulation work:

The obdurancy of these ‘tiny’ barriers presents, at first glance, a puzzle in
human irrationality. [. . .] Why do people persist in using less functional, but
more routine actions when cheaper alternatives are nearby? Are people so
routinized, so rigid in their ability to adapt to change that even such a slight
impediment is too much? [. . .] One way to explain this magnification process is
to understand that in fact two processes of work are occurring simultaneously:
Only one is visible to the traditional analysis of user-at-terminal or user-with-
system. That is the one that concerns keystrokes and firnctionality. The other is
the process of assemblage, the delicate complex weaving togedrer of desktop

resources, organizational routines, running memory of complicated task queues

129

(only a couple of which really concern the terminal or system), and all manner

of articulation work performed invisibly by the user. (p. 386)

Browsing, it appears then, accomplishes some sort of important articulation work for

Lucy that searching does not.

Part of what browsing seems to do for Lucy is to impart a greater sense of
control. As Barreau and Nardi (1995) put it: “users prefer location-based filing because
it more actively engages the mind and body and imparts a greater sense of control” (p.
40). As with Lance’s archiving practices, I would speculate that a significant
contributor to Lucy’s sense that she needs to control her memories lies in the kind of
knowledge that is most valued at Software Unlimited and the ways in which this

knowledge is used to accomplish the work of the organization.

The data I collected from the other inforrrration developers helps clarify this
sense of control by pointing to the enhancement to memory that the engagement of
multiple senses can give. Although I was not able collect data at the same level of
detail from my other participants as I was from Lucy using Morae, most of them
indicated a similar preference for location-based browsing. Certainly, Lance’s principal
archiving practice, his “back of the mind folder” was precisely the kind of carefully-
organized filing structure that encouraged browsing. Similarly, Monica, in answer to
my interview question asking if she found that the process of managing projects grew
more difficult as the project continued, replied that it did not thanks to the careful file
fOlder structure and naming conventions she employs at the beginning — habits that

facilitated browsing:

130

I feel like I do a pretty good job at the beginning trying to set things up — a
structure. We did a really good job I drink of organizing our N drive and I try to
follow that and put documents where they should go. I have multiple files or
folders and piles of paper. I usually get the initial structure for how I want to set

up the documentation, where I’m going to put . . . set up on the N drive.

When I asked a similar question of Peter about how he got up-to-speed when he first
joined the company, he couched his answer in terms of browsing too: “I’d find out that
there’s a folder on the N drive called ‘current software’. And then once I did go there to
look for it, I looked at everything else that was drere too. So I did a little exploration as
I needed to.” Becky, for her part, characterized her success at managing multiple
projects simultaneously as a result of the carefully organized folder structure she
maintains in Outlook: “as I start getting stuff, I usually for every project have a specific
email folder so I can file everything. And if it gets too big I break it down, like
‘meeting notes’ and different pieces of what I have to do. And in my inbox if there’s
anything I have to address, it doesn’t go in a folder so I’ll know I haven’t addressed it

99

yet.

Only Angela appeared to have a different set of preferences, again, apparently,
due in part to her lengthy tenure with the company, and, as with archiving and
reminding in Chapter 4, the ways in which Angela’s preferences differed from everyone
else’s may help illuminate the characteristics of the memory regime that help shape
these preferences. In short, Angela indicated that browsing was not her preferred

method of finding information and she lamented the time and effort required to render

131

information browse-able. For instance, of Outlook email, which Becky and Lucy spend

so much time organizing, Angela asserts:

Typically email is one of the biggest ways drat we communicate, and it’s not the
best way. Things get lost, you know I’ve got 1500 emails right now in my inbox
and 600 of them aren’t read yet. [. . .] Not only just emailing and sending email,
but sorting it, filing it, cleaning it up, deleting it, you know all the stuff that

comes with email. [. . .] This has always been a challenge.

Angela further expressed a degree of pride in her ability to function efficiently without
dedicating much time to organizing memories, as her newer colleagues must do: “other
people, they have file folders and they keep track of things. I just don’t.” In other
words, it appears that Angela simply does not consider the time and effort required to
create and maintain a careful location-based memory system to equal the benefits to be

gained by doing so.

Yet, as I noted in Chapter 4, this is not a sign of carelessness on Angela’s part.
It appears instead that Angela’s disdain for the effort entailed in archiving in order to
enable location-based browsing springs from the hexis that her lengthy tenure with the
company has enabled her to develop. This hexis has made maintaining a detailed
apparatus of external memory unnecessary. In other words, I would posit that Angela’s
memory “locations” are principally the spaces in which she works: the product
interface, the location of her office, and the team of people she is surrounded by.

Consequently, Angela does not need to create and maintain other browsable structures.

132

She does not feel drat she needs the additional level of organization in order to control

information.

I noted in Chapter 4 that the source of the digital aspect of Angela’s hexis lies in
part in her intimate connection with the origins and evolution of her product: “One of
the goals that has been in place since I started with [Software Unlimited] was design a
[. . .] product that Angela can actually use.” But it also lies in Angela’s relationships
with the product team members with whom she has worked for years to build that
product: “I worked with the developers to, for the first time ever, put in workflows and
processes into the systems.” Moreover, these relationships are strengthened by her
close physical proximity to these product team members, which enables her to, in a
sense, absorb information from her physical and social environment without conscious

effort:

From the da y that I got here I was always part of development and I sat with the
development team in their project area and then all of a sudden we started to
grow and they were trying to figure out how to seat everybody [. . .] So at first
they divided us [i.e., the information developers] up [. . .] But we noticed that we
were out of the daily hub of what was happening with the product [. . .] we
weren’t able to glean as much information just from floating fiom what you hear
or being able to just walk next door and ask a question. And it started to show
in the development of the products that we weren’t keeping —- being kept —
abreast of as much as we should have because we weren’t there for all that
spontaneous discussion and talk and brainstorming that takes place. [. . .] Right

now everybody besides myself has their own place in the information

133

development area. I moved into another area. [. . .] I’m part of an experiment
which is different from the rest of the people on my core team, and it’s proving
to be much better. Being able to be with the product team we know that that’s

the best way to go.

Angela’s team occupies this experimental, “Petri dish” area (see Figure 2), which
consists of several two-person offices surrounding a common team meeting area. This
combination of familiarity and proximity enables Angela to find the information she
needs quickly by referencing her colleagues rather than searching and browsing a
database: “So on a continuous basis, if I run across something in the product as I’m
using it, I can either IM somebody or get up out of my chair and holler around the door
[. . .] We have meetings that are right outside everybody’s door [. . .] So I could just step

outside the door.”

The other information developers have, by contrast, learned to rely more on
digital tools, such as instant messaging or 37signals’ Campfire, to maintain team
awareness and coordinate group work. According to Angela, “the other teams are
spread out across four buildings. They need a collective way of gathering thoughts.
We have lots of white boards that have writing on them and, we are kind of the envy of
the rest of the company, we’ve got the huge monitor. We can set it up. We’ve got
movable firmiture for moving configurations and tables.” In comparison, Peter noted
the difficulty in coordinating work on his product teams without interrupting his own

work or intruding on his colleagues:

134

I spent some time this summer actually over there in the other building. And
that was very good when I was working with that team, but at the same time it
made working with other teams harder [. . .] We use Campfire a lot, which is
kind of a team chat room. That happens a lot during the day when anyone has
questions — even when they are in the same building and they have offices next
to each other — they can’t quite talk to each other all the time without getting up.

It’s just a way to ask questions real quick.

Similarly, Lucy informed me that the only person she tended to routinely consult if she
had a question while working was Lance, primarily because his cubicle was adjacent to

her own and she could yell to him without leaving her seat.

The difficulties in coordinating teams using software alone have been
documented in studies from various theoretical perspectives, including joint-cognition
theory (Hayne, 2005; Perry, Fruchter, & Rosenberg, 1999) and activity theory (Swarts,
2005; Spinelli, Perry, & O’Hara, 2005). The evidence from these studies points,
generally, to the conclusion that physical or spatial proximity adds an informational
dimension to quotidian objects and interactions that enhances communication and is
hard to duplicate in software. In other words, the lack of physical proximity adds a
need for an additional level of articulation work just to stay on the same page (to borrow
a metaphor from the age of print). In fact, one of the major purposes of Software
Unlimited’s adoption of the Agile medrodology, with its emphasis on daily face-to-face

team Scrum meetings, was to maintain just this kind of team awareness.

135

However, despite their overall enthusiasm for the brevity and conciseness of the
Serums, the majority of the information developers expressed some reservations about
them, indicating that Serums alone were not adequate to coordinate the teams’ work and
that the Agile methodology had not been uniformly implemented or adhered to by all
the product teams in the organization. Lance, for instance, noted that although he
appreciated the team momentum generated by the Scrum he did not always find it very
useful: “it’s not the most insightful thing in the world and a lot of people don’t have
anything to say really.” Similarly, Lucy noted that, because her team’s Scrurrrs did not
always provide the kind of information about her teammates’ progress and activities
that she needed, she was sometimes unable to complete her own tasks: “nobody ever
says there are impediments -— there are never any impediments [. . .] we do go over stuff
but it just comes out like [. . .] sometimes it’s like somedring should have gotten done,
but no one mentioned that they had worked on it or were going to work on it — it seems

that there’s somedring missing in this tracking.”

In summary, it appeared that the other information developers lacked the sense
of control of their working lives that Angela possessed as a result of her deep and
thorough knowledge of her product, her lengthy relationship with her product team, and
the spatial organization of her office area, which made communicating with her product
team ‘natural.’ I would posit, then, that one possible contributor to the other
information developers’ preference for location-based browsing may be that browsing
gives them a sense of control, both mental and physical, of the memories of the
organization —— it “places” product and other organizational knowledge into a “real”

space where it can be both seen and touched. This corresponds with Haas’ (1996)

136

notion of “tangibility,” which she calls “a complex of both visual and tactile aspects of
interacting with texts” (p. 71). Has notes drat “a computer system or other technology
has high tangibility when users feel that they are in direct contact with the text” and that
this sense of contact is enhanced when users can “touch” the objects on their screens,
such as they can during location-based browsing: “pointing a mouse directly at a word
the writer wants to highlight is more tangible than using arrow and function keys to
select the wor ” (Haas, 1996, p. 71). Interpreted in this light, the prevalence of Lucy’s
“touching” behavior during browsing and reading, then, appears to be yet another
manifestation of the urge to establish control by bringing knowledge into the world of

the senses.

Finally, one other type of gesturing — the imagistic gesturing that Lucy
performed when describing her memory practices during our initial interview — may
suggest another, perhaps even more important function that location-based browsing
serves for the information developers. To describe this function, I will consider in
detail one particular portion of my initial interview with Lucy, a brief excerpt from her
description of her “data dump,” and juxtapose this description with her use of this same

data dump during the second scene of memory work described above.

Before I begin, however, I will, again, preface my discussion by noting drat,
although a single incident performed by a single participant cannot be generalized to the
organization as a whole, I interpret the incident as significant because I recorded, to
varying degrees, the same type of gesticulation among all of my participants when
recounting their interactions with the software interfaces by which they perform their

memory work. Also, I should note that, as I have done at several other points in this

137

dissertation, I have added my interpretative descriptions of Lucy’s hand and body
movements in brackets in the transcription. My descriptions of these movements are, of
course, based on my subjective interpretations of what I believe I was seeing in the
videotapes as I reviewed them multiple times. My descriptions are, therefore,
inherently less objectively reliable than the verbal portions of the transcript. Further, I
have not at this stage in my research attempted to subject my interpretations of drese
movements and gestures to the test of inter-rater reliability. Instead, I have indexed the
movements to individual frames extracted from the video (Figures 10-12), which will
serve, I hope, to substantiate my interpretations to some degree. The transcript is as

follows:

I have like a Sprint folder [Figure 10, drawing the folder on the tabletop with her
fingers as she speaks] and dren in there I have different things. But I have one
folder called ‘Features’ [Figure 11, drawing this folder on the tabletop] and in
the folder ‘Features’ I have a different folder for each [Figure 12, drawing these
folders] of the things that we are working on: ‘Power . . .’ [interrupting self]
‘PPT,’ um ‘Undo,’ um [looking off into distance for several seconds] ‘Camera
Recording.’ Just everydring that we are working on and then I have those

documents in there.

I will begin by calling attention to the two memory “lapses” that are noticeable in this
transcript. First, Lucy starts to describe one feature folder as “PowerPoint” but
interrupts herself and changes the folder name “PPT.” Second, in a classic example of a

movement indicating a “speech failure” (McNeill, 1992 , p. 77), Lucy pauses and looks

138

off into the distance while trying to recall the name of another subfolder, which she

ends up calling “Camera Recording.”

Figure 10: Lucy “draws” the Sprint folder

 

Figure 11: Lucy “draws” the Features folder

 

Figure 12: Lucy “draws” the subfolders

   
 

   

. fiw‘qrfi‘cfixflflfi!

139

 

Next the form of the gestures appears significant. Lucy informed me that she is
right-handed and that she only uses her right hand when manipulating her computer
mouse, which was confirmed by my observations. Yet generally, with the exception of
the imagistic gesturing in which she describes using her computer mouse such as those
occurring in the transcript above, Lucy tended to be ambidextrous when gesticulating.
For instance, Figure 13 provides an example of Lucy gesturing with her left hand when
describing the action of replacing text on a screenshot. The significance, then, is that
the memory drat Lucy recounts in the transcript lives not just in her mental image of her
folder structure but also in the kinesthetic memory of her right hand. This finds
additional confirmation by considering the motion of Lucy’s hand that can be perceived
by observing Figures 10-12 in order. As she describes the folder structure, Lucy moves

her hand progressively to her right in the same pattern she follows as she navigates

Figure 13: Left-handed gesture

 

the folder and subfolders with her mouse cursor in Windows Explorer. That is, when
compared with the two views of Lucy’s data dump folder in Windows Explorer in
Figure 14, which are excerpted from the Morae recordings of Scene 2 when Lucy was

attempting to find her working document, Lucy’s hand motion appears to follow the

140

same path. In short, I would argue that Lucy’s gestures are, as McNeill (1992) predicts,
revealing her “memory image” and “particular point of view” of one of her typical
activities, but that they are also showing, as Aristotle predicts, that this memory image

is composed of both visual and kinesthetic components (p. 13).

Figure 14: Two views of Lucy’s data dump folder structure

 

 

 

 

 

 

 

+ .3 $er 41002 - ‘3 Sprint 21001
+ «J L I _) Features
" __3 Sprint 21% k ,4} WNW
" JCliDbOB'd _) MiltLMontor
+ .3 1.99% .3 PPT_Export
._J Mukijtlonitor + «J Task—Savoy;
+ .3 Ports _) Veriablejpeedjleybeck
t .31 PPT + -3 Smportimmaterids
+ .3 Snapshots - -3 Sprint 21002
t .31 Two_Camera + 3 PPT _Export
.3Undo + .3: "7" ““3
.+ .3 Sprint 30005 +, ____3 Sprint 21004
+ __) Sprint 30006 + J Sprint 30005

 

 

 

Further, comparing Lucy’s gestures with the momentary lapses that I noted in
the transcript above — her mid-word self-correction from “Power-” to “PPT” and her
pause to remember another subfolder — it appears that the gestures are serving an
important function for Lucy: they are not merely displaying her images, they are
actually helping her to remember them. As McNeill’s more recent work on gestures
puts it: “gesture can materialize the spatial structure of the problem [. . .] Spatial
structure is a mnemonic, especially when extemalized. The linkage of memory to
spatial structure was known in ancient times. Roman orators made use of the layout of

space around them to recall the points of a speech (2005, p. 148). Lucy’s picturable and

141

touchable folder spaces, then, are serving a multi-sensory mnemonic function for her as
she attempts to remember and articulate for my benefit their contents. This is especially
apparent in Lucy’s shift of phrasing from “Power-” to “PPT.” If her attention was
focused primarily on describing for me the folder’s contents widr maximrun clarity, it
seems to me that Lucy would have finished the word “PowerPoint” and left it at drat,
but, as her pausing and frequent “urns” indicate, she was actually focusing a great deal
(perhaps the majority) of her attention on just remembering those contents. So, as she
scanned her mental image widr the eyes of her mind and touched it with the memory of
her hand, she articulated the contents as she saw them — “PPT” instead of

“PowerPoint,” as Figure 14 reveals.

This incident highlights, then, the other major function that location-based
browsing appears to serve for the information developers: it is a method of locating,
stabilizing, and embodying product knowledge, the knowledge most highly valued by
the memory regime at Software Unlimited. In particular, Lucy’s data dump folder
structure, which collects and organizes products into Sprints and Features, appears to be
an externalization of her standing wave of memory about the evolving product
interface. That is, Lucy “consigns” the product to her memory in both of Derrida’s
senses of that word: by writing about it, she transforms the product into the signs with
which she is most comfortable, text (e. g., she says “I just write”), and by placing these
texts into her data dump she gathers them together into her own memory places. She
then routinely “walks” these memory places through the activity of browsing, a litany-

like process by which she imprints the knowledge into her visual and bodily memory.

142

Conclusion

Even when her elaborate filing scheme does not appear to be working correctly
— such as in the breakdowns recounted above — Lucy’s memory systems are still
performing important articulation work for her by helping her learn her product better.
What ultimately appears to be most important about this memory work is less the
written products with which she populates her archive than the processes of archiving
and finding themselves, which implant knowledge more firmly into her memory. In
other words, the activities of memory are contributing to the development of Lucy’s
hexis. A point that Winsor (2003) makes about the employees of the engineering center
that she studied seems to apply equally well here: “The tools they used were not
incidental to the communication and knowledge generation in which they were able to
engage. Rather, knowledge about how to analyze and classify knowledge had become

stabilized and accessible in the material forrrr of the tool” (p. 136).

Perhaps the most interesting finding of this chapter, though, lies in the fact that
the memory regime compelled Lucy to expend such time and effort in the first place.
That is, where the scenes of Lance’s memory work demonstrated that the effort of
stabilizing less-valued forms of knowledge, like user knowledge, could also at times
yield valuable results, such as changes to the product interface, those of Lucy show her,
I would argue, forced to spend undue amounts of time engaged in non-creative, non-
inventional labor (the equivalent in some sense of rote memorization) just to fulfill her
most basic and lowest-status job duty: writing documentation. As was the case with

Lance, the obscuring of the effects of the memory regime by the quotidian practices

143

necessary just to cope with these effects appears to be the principal barrier to reforming

the memory regime: it is hard to change what you are too busy to see.

144

CHAPTER 6: CONCLUSION

Introduction

The original aim of this study was to explore the role of embodied contexts in
mediating the memory-related activities of invention. I anticipated that such an
exploratory study would increase our understanding of the digital, physical, and social
interfaces and infrastructures for doing memory work in organizations, thereby giving
us a basis for arguing for changes to drose interfaces, spaces, and organizations. To
accomplish this purpose, the study asked the following research questions, which I
envisioned as a first step towards articulating something that might be understood as a

rhetorical practice of memory:

0 Do writers use the affordances of their embodied contexts to support memory-

related demands of composing?

a Where/when in the composing process do embodied-contexts seem to be most

important?

0 How, if at all, do embodied contexts aid composers’ memory work during

invention?

Based on the analyses in Chapters 4 and 5, I would tentatively answer “yes” to the first
question and add that, as I will show below, the process of explaining why I answered
in this way will probably end up answering the second and drird questions as well.
During the course of my research, however, I began to discern a possibly more

important theme running drrough my findings. This theme may help illuminate and

145

expand upon Bowker’s concept of the memory regime and enrich our understanding of
the memory work that technical communicators perform in their workplaces in ways
that go beyond the original aims of my research. I will address each of these sets of
findings — the findings related to answering my research questions and those related to

expanding Bowker’s concept of the memory regime — in turn below.

What the Study Revealed about an Embodied Rhetoric of Memory

To begin considering how my findings help answer my research questions, it is
worth revisiting several aspects of the conceptual framework that I articulated in
Chapters 1 and 2. I will start with the three-part formula for understanding memory
work as a rhetorical practice of invention in knowledge-intensive workplaces that I
formulated in Chapter 1. First, based on my reading of contemporary workplace
theorists like Reich combined with my understanding of memory practices in the
classical rhetorical tradition, I posited that responsiveness to kairos, or the ability to
retrieve and use memories in a timely and appropriate manner, was more critical as an
indicator of successful memory work than raw storage capacity. I tied this necessary
responsiveness to kairos to the perceived need to avoid information overload, a problem
which we all seem to face these days. Second, I noted that affordances for offloading
the cognitive burden of performing memory work to infrastructures or tools or to
coworkers was an important component of successful memory work. Third, and
perhaps most germane to my research questions, I argued that the ability to ‘touch and
feel’ information, to bring it into the world of multiple embodied senses was crucial to
making memories usable (e.g., responsive to kairos, offloadable, and manipulable). I

derived this last insight principally from my reading of the rhetorical memory

146

techniques of the ars memoria, particularly the walking mnemonic, which seemed to me
to be an ingenious and effective, though laborious, method of responding to rhetorical
situations, of offloading information to a “software” tool, and of using bodily
perceptions to do memory — in other words, of thinking on (or perhaps with) one’s feet.
I further held that any interface or other infrastructure that did not take into account
these three aspects of memory work in invention would to some degree impede or make
more difficult writers’ abilities to perform symbolic analytic work for their

organizations.

So what about my research questions? Did embodied contexts appear to play a
role in the information developers’ memory practices in this particular memory regime?
Did they, in other words, play a role in helping the information developers perform their
articulation work by, for instance, making product or user knowledge more responsive
to kairos, more shareable, or more personalized? And, if so, where, when, and how did
they do so? To answer the first question, I will revisit the idea of the “standing wave of
memory,” which I articulated in Chapter 2 as a sort of ideal condition of memory in
which we are able to achieve a working version of the past to enable action in the
present. In short, it appeared that the newer employees at Software Unlimited had to
perform a great deal of extra work to try to stabilize this standing wave than did the
more senior ones. That is, Lance and Lucy appeared to put a great deal more effort into
their memory practices than did Angela. From this, I arrived at the notion that
achieving this standing wave requires somedring like a hexis, an assured facility of mind

and body that renders its practitioner capable of performing both input (storage) and

147

output (finding) activities of memory work naturally, “just by floating,” as Angela puts

it.

In addition, there did indeed appear to be a physical, embodied component both
to Angela’s hexis and to Lance and Lucy’s attempts to acquire a similar degree of
mastery. Angela, for instance, seemed to attribute much of her ability to “just know”
drings to the fact that she worked in physical proximity to the software developers for
such a long time. Similarly, it seemed to me that part of the motivation behind both
Lance and Lucy’s labor-intensive memory practices was to inculcate in themselves a
mastery of product knowledge through mental and physical training: Lance dedicated
considerable time preparing elaborate and visually-complex documents to take to
meetings because they could serve as cues and objects of attention as he articulated his
arguments in words and gestures, while Lucy employed her elaborate and redundant
archiving system at least partially as a spur to acquire product knowledge drrough the
repetitive activity of browsing through her complex file system. As I noted at the
beginning of the chapter, the answer to the first research question, then, appears to be
that embodied contexts did play a beneficial role in invention by helping the
information developers respond to the kairos of Scrumming, to offload memories into

reminders of various sorts, and to personalize memories through repetition and habit.

This leads to the final questions concerning where, when, and how embodied
contexts played the biggest role in invention. It appears from my data that the answer to
all three of these questions is that embodied contexts were most helpful in the times and
places in which my participants had to communicate, transport, or rearrange knowledge

in some way. In other words, embodied contexts were most important when memory

148

 

moved because the affordances perceptible to their senses, principally sight and touch,
helped my participants to transform raw information into useful memories: Lance
rendered user knowledge concrete using visuals and gestures and drrough mimesis;
Lucy rendered product knowledge memorable and understandable by using browsing to
inculcate a habit of both hand and mind. Moreover, this appeared to be true of both
storing and finding activities in a scene of memory work: the transformation and
movement of memories into the archive drrough consignment appeared to benefit from
multisensory affordances as much as the recall of those memories from the archive
during use. That is, as Monica and others indicated, the act of setting up the archival
structure of a browse-able file folder system at the beginning of a project served a

critical role in all subsequent work on that project.

What the Study Revealed about the Memory Regime: A Final Scene

Although my theoretical framework did appear to answer my research questions
in productive ways, on the very first day that I conducted observations at the research
site in September 2007, I observed an incident of memory work (the prototype for what
I would later came to define as a “scene”) for which my theoretical framework proved
inadequate to explain. This scene stuck with me and resonated throughout my study
because it raised issues and introduced themes drat I was to see recur repeatedly as I
followed my participants over the next six months. The recurrence of these themes and
issues and the inadequacy of my original theoretical focus to explain them eventually
led me to reconsider Bowker’s memory regime as a concept reconciling the
bidirectional influence of local workplace cultures and memory practices. Adapting the

method of explicating a scene that I employed in Chapters 4 and 5, I will first describe

149

the events of this scene before discussing the issues it raises in a way which, I believe,
will clarify the role of the memory regime on Angela’s hexis and on the memory

practices by which Lance and Lucy attempt to achieve their own expertise.

Overview of the Scene

I accompanied the Information Development team to lunch on the afternoon of
September 13, 2007. After lunch, Angela asked me if I would like to observe her
weekly meeting with her “partner” from the training team, “Carl.” As. the information
developers and I walked into the Information Development wing, Becky added that
Angela would be a great person to observe because she knew so much about the
company and, in fact, was one of the first employees hired, possessing an employee
number in the low 205 (employee numbers originally reflected the person’s hire number
with the company, a practice that I learned had subsequently been discontinued as the

company grew).

Angela informed me that she and Carl would be meeting to discuss the number
and content of the training tutorial videos Carl would be creating for the next release of
their product. I later learned that each member of the Information Development team
had recently been partnered with the member of the training team assigned to the sarrre
product in order to coordinate the training and documentation efforts, to ensure
consistency across documentation, and to avoid unnecessary duplication of effort on

each product release.

150

 

Synopsis of the Action of the Scene

The meeting took place around a small table in the petri dish area of the
Information Development wing (see Figure 15). As I did throughout the study, I
attempted to observe the meeting without intrusion by positioning myself slightly
behind my participants. When we arrived, Carl had already set up his laptop on one of
the mobile tables in the petri dish. He had positioned his laptop so that Angela and I
could see the screen as they talked and had already opened on it a “mind map” that he
had created using the software Mindjet Mind Manager, an application that, I later
learned, the training team regularly used as a tool for visually brainstorming ideas but
that the Information Development team did not possess. The mind map that Carl had
created showed his tasks and ideas for the training videos that he planned to create for
the upcoming release of his and Angela’s product. Later in the meeting, he also opened
the current build of the product itself in order to deploy the most recent version of the

online help that Angela was creating for the release.

The discussion concerned how many training videos Carl needed to create for
the upcoming release. Carl expressed his concern that there should perhaps not be a
one-to-one correspondence between the topics in the online help and the training
videos. He reasoned that the 30 topics in the online help would translate into far too

many videos for him to create in the allotted time.

151

Figure 15: Angela and Carl’s meeting in the petri dish

 

 

Carl’s
Laptop

 

 

 

 

 
 

Angela

So, Carl’s primary purpose in this meeting was to learn which user tasks were most
important so that he could produce videos only for those tasks. As part of this
discussion however, Carl volunteered some constructive criticism of the online help in
its current form, noting politely that the current visual format looked good but that he
believed it offered too many options in perhaps the wrong order for users to quickly
find what they needed. He contributed one further piece of criticism by asking Angela
if it would not be useful to incorporate small icons and pictures of buttons from the

product user interface in the text of the online help itself.

As Carl made these suggestions, Angela’s body language changed — she sat up
straighter, seemingly less-relaxed than before, and when she answered Carl’s points, she
began to rather stridently jab her pen at his laptop screen to underscore the points she
was making. She rebutted Carl’s recommendations by referring to her own experience
and research, which, she said, showed that users preferred the online help in its current
form. Perhaps realizing that he had offended Angela and overstepped the purposes of

the meeting in offering these critiques, Carl said he would defer to her superior

152

experience and knowledge and he added, rather meekly, that he had only been with the

company for a short time.

During this discussion, Carl’s laptop screen had been the primary focus of
attention, as Carl alternated between the online help and his mind map. As I mentioned
above, while responding to Carl’s critiques, Angela began to use her pen to make
deictic gestures frequently to underscore the points she was making in speech. She
seemed to be particularly emphatic in these pointing gestures when referencing Carl’s
mind map. This behavior was so pronounced that I noted in my research journal later
that day that it appeared as if Angela were attempting in some way to puncture the
invisible barrier between herself and Carl’s thoughts as they appeared on the screen as if
this would enable her to get inside his head and better understand his reasoning. At one
point in this, Angela informed Carl that she did not have the Mind Manager software
and asked if he would email her a screen capture of his mind map to help her prioritize
her own tasks. Angela then added that she would block time for these efforts in
Microsoft’s Team System software, a team awareness application drat enables software
team members to view each others’ schedules in order to coordinate work efforts and
arrange meetings. She asked Carl if he used Team System, but he indicated that he did
not use it much, again mentioning that he had only been employed at the company for
six months and adding that he was still adapting to the way things were done there. He
further added drat he had not experienced a similar degree concern for team awareness
in his previous job and had been surprised by this aspect of the work environment at

Software Unlimited. At this, the meeting concluded and Angela returned to her office.

153

Analysis of the Scene

Angela’s agitation at Carl’s critique was the most interesting thing about this
scene. At a number of points later in the study, I attempted to learn from Angela what it
was about this meeting and about Mind Manager that had so agitated her, but in each
instance I found her reluctant to talk about the incident. In fact, although I had initially
found Angela to be the most enthusiastic participant in the study — a fact demonstrated
by her request that I attend her meeting with Carl on the first day of the study — after
this meeting she appeared reluctant to participate in other observations sessions and I

began to focus on more willing participants like Lucy and Lance.

At one level, of course, the reasoning behind Angela’s change in behavior seems
obvious — she was probably embarrassed drat I had witnessed her handiwork being
critiqued, even though she had won the argument. At another level and in light of
Angela’s stature with the company and depth of knowledge about the company’s
products, her behavior appeared somewhat puzzling: why did she care so much if a new
employee who was ignorant of the drinking that had gone into her designs should
attempt to offer his advice? She clearly knew more than he did and handily won the
argument by invoking her superior knowledge and understanding of the company and
its users — “why it is the way it is,” as Lucy put it. So, why did Carl’s criticism appear

to sting so?

Although this question cannot be fully answered with reference to the data I
have collected — the explanation for Angela’s behavior may have nothing to do with the
events of this meeting, perhaps she simply got involved in more pressing issues than

accommodating my research —- Bowker’s memory regime offers one possible

154

explanation, an explanation that does happen to jibe with the rest of my data concerning
the status of various memory practices within this organization and the source of

Angela’s own expertise.

To arrive at this explanation, it will perhaps be helpfirl for me to first recap
Bowker’s rather incomplete articulation of the concept of a memory regime and then to
explain my interpretation and expansion of the concept with reference to the memory
practices that I observed at Software Unlimited. First, as I have already mentioned, I
follow Bowker in adopting the notion of the memory regime as a concept describing the
role of culture(s) in shaping memory interfaces, infrastructures, and practices. Bowker
(2005) says of memory regimes that they “articulate technologies and practices into
relatively historically constant sets of memory practices that permit both the creation of
a continuous, useful past and the transmission sub rosa of information, stories, and
practices from our wild, discontinuous, ever-changing past,” which is as apt an
explanation for the effects of culture on memory and the effects of memory on culture

as I have found (p. 9).

I found the concept of memory regime, then, to be a good way of talking about
the memory work of cultures, in this case of a particular workplace culture: “culture is
the work of selection: the selecting, challenging, arranging, and living of these received
artifacts and ideas in everyday life” (Slack & Wise, 2005, p. 4). Memory regimes do
the cultural work of selecting which memory practices are valued and how those
practices are, for lack of a better word, practiced. Further, one of the purposes of a

memory practice or set of memory practices is in some sense to obscure the fact of their

155

 

own existence — that is, like any other technology, memory technology works best when

we do not see it.

Aside from introducing the concept, however, Bowker had not done much to
develop the idea in the empirical portion of his research, so it was my task to figure out
how to use it to interpret the patterns I was seeing in my data. In short, as my accounts
of the scenes with Lucy, Lance, and Angela have argued, I kept finding instances in
which my participants would, in one way or another, allude to some sort of gnostic yet
important type of knowledge in the organization that seemed to have a large effect on
how the work of the organization got done. I reasoned that the unspoken and apparently
invisible rules governing this knowledge were the primary indicators and delineators of
the memory regime of the organization, and I saw this regime determining three drings:
a) which types of knowledge were considered most important; b) the forms in which this

knowledge was preserved; and, perhaps most importantly, c) who had access to it.

First, as for which types of knowledge were most valued by the memory regime
at Software Unlimited, it was Lucy who had provided the first clue that it was
knowledge of the company’s products that was most valued — knowledge both about the
features and functions of the current versions of the products and about the arguments
and rationales lying behind the design decisions that had been made during the
evolutions of those products. I found it highly significant that Lucy introduced the
topic of the back reasons by interrupting herself during a discussion of the data dump
system to which she consigns product knowledge: “There’s probably way more
information in it than I’ll ever put out. And I think drat one thing that I’ve noticed here

is all of the reasons —- back reasons — [Angela] has been huge.” The interviews and

156

behavior of the other information developers seemed to confirm my supposition: Lance
struggled to make a case for user needs against Wallace’ superior knowledge of the
product history; Peter informed me that he knew he needed to pay extra attention
whenever the software developers appeared to be having problems designing a certain
piece of the product; and Angela indicated that she “just knew” the “historical stuff”
precisely because of her long association with her product — she had been, quite

literally, her product’s first user.

As for the forms in which this product knowledge was preserved, I have already
discussed the fact that the product interfaces provided an ever-present but in large part
inscrutable record of dreir own evolution. Interpreted in light of the cultural work of the
memory regime, Angela’s hexis demonstrates the other, crucial, form of memory which
makes the other memories of the organization readable: narrative. In short, by virtue of
her long tenure with the company but also by virtue of all the other advantages she has
accrued — her physical proximity to her product team, her long-standing relationships
with all strata of her fellow employees, etc. — Angela has learned more of the back
stories, more of the narratives of the products’ evolutions than any of the other
information developers. As the scene with Carl demonstrates, it is Angela’s access to
these stories, to this knowledge that in large part appears to be the source of her

influence and authority.

In addition, I found it significant that Angela, more than any other study
participant, tended to answer the initial interview questions with narratives of specific
events from her work experience, particularly with stories that focused on her

relationship with her product. This was especially noticeable when the questions called

157

 

for relatively straightforward expository answers, such as in the following quotation
excerpted from Angela’s response to the question “what software tools do you use for

composing?”

When I came here, since it was just me, I used Word. [. . .] One of the goals that
has been in place since I started with [the company] was design a [product] that
[Angela] can actually use. When I got here and started documenting and
looking at [the product] there were lots of drings that were not in [it] so I could
use it to do all of my [work] And the workflows may have been there for a lot of
people but they weren’t right for technical writers. So we came up with the add-
ins at that point and we developed those. [. . .] So that was Phase 2 of if [Angela]
can use it. And then we came up with the output from right widrin [the product]
[. . .] So, it has really been a goal for the past five years to make [the product] as

robust as possible.

Again, what stands out about this answer, typical of Angela’s answers to many of the
questions in the interview, is its narrative quality, the way in which the memories of the
important decisions of the organization are preserved as stories, stories intricately
connected to Angela’s own history with the company. Along with Lucy’s revelation of
the importance of the “back reasons” and Becky’s assertion that “we are really lucky
because she [Angela] has got so much background and understands sort of how we got
to where we were with some of these choices and different drings that were tried,” the
form that Angela’s answers take reveals perhaps the ultimate source of her hexis — she
knows the stories of the organization. If we accept Lyotard’s (1999) assertion that

“what is transmitted through [. . .] narratives is the set of pragmatic rules drat constitutes

158

 

the social bond,” the true source of Angela’s stature and authority widr the company

stands revealed (p. 21).

Finally, there is the issue of access. As I have attempted to show, the memories
of Software Unlimited’s products appeared to be the most important memories and the
forms in which these memories were stored and most effectively retrieved was in stories
or narratives of the design decisions related to the creation of these products, but these
aspects of the memory regime were problematic because, of course, access to these
back stories depended on an employee’s longevity with the company and the right set of
skills or team affiliations to acquire them, and thus they were not evenly distributed
among the employees. That is, although product knowledge was always present and
visible in the software products themselves, principally in the user interfaces, this
knowledge was not always interpretable or understandable to those whose newness or
status as non-software developers meant they did not know the contexts and
circumstances in which these memories were created. The memory practices of the
newer information developers, then, appeared often to be attempts to “consign”
knowledge in order to make it usable (i.e., contextualized and findable): “In an archive,
there should not be any absolute dissociation any heterogeneity or secret” (Derrida,

1995,p.3)

While knowledge of the products appeared to be the most important indicator of
status and served as the ultimate warrant for many of the arguments about the way the
work of the organization would get done, another type of knowledge that I encountered
— user knowledge — helped to make these uneven power relations visible and perhaps

helps to explain Angela’s discomfiture at Carl’s critique of her online help. As I noted

159

in Chapter 4, in comparison to product knowledge, knowledge about users appeared to
be relatively homeless and less important, and this seemed to have a large effect on the
status of the Information Development team members who were, in some sense, the
principal custodians of that type of knowledge. Since other teams did not appear to
value user knowledge as highly as the information developers and since it did not have
a visible habitation as product knowledge did, the newer information developers had to
engage in a variety of extra articulation work just to make this knowledge visible, to

grant it the status of topoi in warranting their arguments for change.

The exception to this tendency to engage in laborious extra memory work was,
as I have already shown, Angela, who was able to win her argument widr Carl handily
just by referencing her superior knowledge of the back story lying behind her product
and its online help. In fact, the scene between Angela and Carl demonstrates several
parallels to the encounter between Lance and Wallace in which Lance’s attempt to
argue for user needs was stymied by Wallace’s appeal to the back story. This, then,
offers one possible explanation for Angela’s discomfiture at Carl’s critique: despite the
much touted role of the information developers as the principal user advocates in the
organization, in this instance Carl was adopting the perspective of the user to critique an
information developer’s work. And, to counter his argument, Angela, perhaps
unthinkingly, employed the practice sanctioned by the memory regime: she invoked the
back reasons. I would speculate that such a reversal of roles may have left bad taste in

her mouth.

In sum, then, over the course of my data collection and analysis, the concept of

memory regime proved quite important as a lens through which I interpreted memory

160

 

practices as cultural practices at Software Unlimited. Like culture, a memory regime is
instantiated at multiple levels and both influences and is itself enacted by practices, each
of which forms “a sort of mini-culture that binds the community together” (Wenger,
McDerrnott, & Snyder, 2002, p. 39). That is, in part the memory regime at Software
Unlimited was derived from the larger culture — the preference for Microsoft operating
systems utilizing the standard file folder metaphor, the privileging of technical
knowledge like product knowledge over knowledge about people, like user knowledge,
etc. — but it was also profoundly influenced by local conditions, local power relations,
and the local history of the organizations and its products, the narratives that bind the
organizational culture. Together, these factors guide the memory practices of the
organization and determined its topoi: what counts as knowledge, how this knowledge
is stored, how and when it can be used in arguments, which retrieval practices are
appropriate in a given situation, and, by extension, who can access and use this

knowledge.

Implications for Practicing and Teaching Technical Communication

At this point, I perceive two main implications for this research: one for teaching
and the other for practicing technical communication. First, for teaching, my analysis
seems to indicate that, because the acquisition of an “assured facility” in technical
communication appears to be a product of both physical and mental training, the
“cognitive apprenticeship” model of education advocated by Brown, Collins, and
Duguid (1989) would be the most effective (p. 39). It seems to me that such a model
would entail longer internships than many technical communication degree programs

currently offer — certainly it would emphasize internships lasting longer than a single

161

summer. The information developers seemed to recognize that such an experience
would be helpful because, even though all but one of them had majored in technical
communication in college and several had held intemships as part of their training, they
all indicated that the most important part of their education had only begun once they
were on the job. Again, Lucy said it best: “drere’s a lot of practical stuff I learned in
college -— like really how to work with technical things [but] the practical stuff to do my

job I’ve learned working.”

As for the implication for the practice of technical communication, I would
argue that this study, particularly as it illuminates the concept of a memory regime,
contributes to our recognition that organizations and organizational memories are
rhetorical constructions drat are “written” in interfaces, spaces, and databases, and,
therefore, that “we can rewrite [them] through rhetorical action” (Porter, et a1, 2000, p.
613). The most obvious example of such rhetorical action aimed at regime change drat
I observed was, of course, Lance’s tireless efforts to substantiate the user perspective as
an area of knowledge equivalent in importance to product knowledge. He was, in short,
attempting to give stories about users equal status to stories about products as

transmitters of the company’s memories.

However, Lucy’s example also contributes insights by highlighting one
particularly deleterious influence of the wider memory regime on a local context: the
reliance on the file folder metaphor as the dominant metaphor for location-based
browsing. As I argued in Chapter 5, Lucy showed a definite preference for location-
based browsing, yet seemingly as often as not file folder browsing led her astray or took

longer than she appeared to anticipate. In other words, Lucy’s example adds weight to

162

the idea drat metaphors are helpful as a way of making knowledge concrete, but it also
seems to confirm Selfe and Selfe’s (1994) contention that an overreliance on hierarchic
metaphors like the file folder may not be the best approach and may also preclude other

ways of knowing or remembering.

Finally, the example of Angela’s hexis, developed drrough her lengthy social
and physical proximity to her teammates and sustained by her authority as a perceived
keeper of the stories of the organization, points to anodrer text that may need to be
rewritten. As Johnson-Eilola (2005) points out, “it is easy to miss the fact that
symbolic-analytic work requires a much larger and more complex information space
than can be supported by the computer alone” (p. 73). At Software Unlimited, stories
appear to play an even more important role in the organization’s memory than
information held in the electronic information spaces of product interfaces and network
file folders. Rendering these stories accessible and interpretable may be the most

important memory work that technical communicators can do in their workplaces.

163

APPENDIX: INITIAL INTERVIEW QUESTIONS

Questions about Participant Background

What is your current job title?

What are your job duties?

How long have you worked as a technical writer?

How long have you worked as a technical writer at Software Unlimited?

Questions about Composing and Writing Practices

When you are assigned to a project, what sorts of research do you perform in
order to prepare yourself to work on that project? Some examples of this type of
activity might include reading the project design documents or meeting
inforrrrally with other members of the project team to discuss the project?

What resources do you typically consult when conducting this research? For
example, do you search for information on the company file server or the
intemet, do you consult colleagues?

How do you track and organize the information you learned through this
research? For example, how and where do you keep notes and other products of
your research?

Where do you typically compose?

When do you typically compose?

How do you prepare your workspace (both your computer desktop and your

physical surroundings) before beginning a writing task? Examples of this type

164

10.

11.

of activity might include creating computer file folders for the writing
deliverables or creating a physical file folder for hardcopies of project
docrunents.

What software tools do you use when composing?

In addition to the project deliverables, do you create supplementary documents
or files for yourself to help you manage information or track your goals as you
compose? An example of a supplementary document might be a working
outline to remind you of your place in the composing process?

Do you use any supplementary software tools to support your composing? For
example, you might use a text-editor like Windows notepad to create quick
reminders for yourself as you write.

Once you have begun writing, do you take many breaks or do you work straight
drrough until you come to a logical stopping point? If you take breaks, what
activities do you typically perform during them? For example, do you talk to
colleagues, visit websites, or go to the break room?

How often do communicate to your colleagues about the project and what
methods do you use for this communication? For example, do you meet with

them face-to-face, do you instant-message drem, do you email them?

Questions about Project Management Practices

What methods and tools does your employer use to help you track project due
dates? For example, what software (e. g. MS Project) or hardware (e. g. a daily

planner) do you use?

165

Do you have any personal methods for managing projects that supplement those
your employer provides? For example, do you post sticky notes for yourself?
If so, is there a system behind these personal methods (e. g. placing the most
important sticky note on your monitor and the rest on your bulletin board)?

Do you find that, as a project continues, the project management aspects of the
writing process grow easier or more difficult? In other words, does managing
the various deliverables and documents you and others create for a project
become more or less time-consuming as the project continues? Please explain

why.

166

REFERENCES

Ackerman, M. S., & Halverson, C. (1998). Considering an organization’s memory.
Proceedings of the I 998 ACM Conference on Computer Supported Cooperative
Work, Seattle, WA, November 1998, 39-48.

Ackerman, M.S., & Halverson, C. (2000). Reexamining organizational memory.
Communications of the ACM, 43(1), 59-64.

Allen, V. (1993). The faculty of memory. In J. F. Reynolds (Ed.), Rhetorical memory
and delivery: Classical concepts for contemporary composition and
communication (pp. 45-63). Hillsdale, NJ: Lawrence Erlbaum Associates.

Aristotle. (1952). On memory and reminiscence. In The works of Aristotle volume 1
(W.D. Ross, Trans.) (pp. 690-695). In R. M. Hutchins (Ed.), Great books of the
western world (Vol. 8). Chicago: Encyclopedia Britannica, Inc.

Barreau, D., & Nardi, B. (1995). Finding and reminding: File organization from the
desktop. SIGCHI Bulletin, 27(3), 39-43.

Beaudouin—Lafon, M. (2004). Improving interaction: Designing interaction, not
interfaces. Proceedings of the 2004 ACM A VI Conference, May 25-28, 2004,
Gallipoli, Italy, 15-22.

Besser, H. (2002, June 3). The next stage: Moving from isolated digital collections to

interoperable digital libraries. First Monday, 7(6). Retrieved January 17, 2007,
from http://www.firstrnonday.org/issues/issue7__6/besser/index.html

Bizjak, P. (2000). Mankind’s memory managers: A new paradigm of library science.
Library Philosophy and Practice, 2(2). Retrieved January 17, 2007, from
http://www.webpages.uidaho.edu/~mbolin/bizjak.html

Bowker, G. C. (2005). Memory practices in the sciences. Cambridge: MIT Press.

Bowker, G. C., & Star, S. L. (1999). Sorting things out: Classification and its
consequences. Cambridge, MA: MIT Press.

Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of
learning. Educational Researcher, 18(1), 32-42.

Brown, J. S., & Duguid, P. (2002). The social life of information. Boston: Harvard
Business School Press.

Burke, K. (1969). A grammar of motives. Berkeley: University of California Press.

167

Carruthers, M. J. (1990). The book of memory: A study of memory in medieval culture.
New York, Cambridge UP.

Cicero, M. T. (1942). De oratore: Book II. (E. W. Sutton & H. Rackharn, Trans).
Cambridge: Harvard UP for the Loeb Classical Library.

Clancey, W. J. (1997). Situated cognition: On human knowledge and computer
representations. New York: Cambridge UP.

Coleman, A., & Oxnam, M. (2002). Interactional digital libraries: Introduction to a
special issue on interactivity in digital libraries. Journal of Digital Information,
2(4). Retrieved July 24, 2007, from http://joumals.tdl.org/jodi/article/view/jodi-
56/52

Colomb, G. G., & Williams, J. M. (1985). Perceiving structure in professional prose: A
multiply determined experience. In L. Odell & D. Goswami (Eds.), Writing in
nonacademic settings (pp. 87-128). New York: The Guilford Press.

Corbett, E. P. J ., & Connors, R. J. (1999). Classical rhetoric for the modern student
(4‘'1 ed.). New York: Oxford UP.

Crowley, S. (1990). The methodical memory: Invention in current-traditional rhetoric.
Carbondale: Southern Illinois UP.

Crowley, S., & Hawhee, D. (2004). Ancient rhetorics for contemporary students
(3rd ed). New York: Pearson.

Czerwinski, M., & Horvitz, E. (2002). Memory for daily computing events.
Proceedings of HCI 2002: Sixteenth British HCI Group Annual Conference,
London, England, September 2002.

Delmas, B. (2001). Archival science facing the information society. Archival Science, 1,
25-3 7.

Denzin, N. K., & Lincoln, Y. S. (2003). Introduction: The discipline and practice of
qualitative research. In N. K. Denzin & Y. S. Lincoln (Eds.), Strategies of
qualitative inquiry (2"d ed.) (pp. 1-45). Thousand Oaks, CA: Sage Publications.

Derrida, J. (1998). Archive fever: A Freudian impression. Chicago: University of
Chicago Press.

De Chiara, R., Erra, U., & Scarano, V. (2004). A visual adaptive interface to file
systems. Proceedings of the 2004 ACM A VI Conference, May 25-28, 2004,
Gallipoli, Italy, 366-369.

168

 

Dourish, P. (2004). What we talk about when we talk about context. Personal
Ubiquitous Computing, 8: 19-30.

Engestrom Y., & Middleton D. (1996). Introduction: Studying work as mindful
practice. In Engestrom Y. & Middleton D. (Eds.), Cognition and communication
at work (pp. 1-14). New York: Cambridge UP.

Fast, K. V., & Sedig, K. (2005). The INVENT framework: Examining the role of
information visualization in the reconceptualization of digital libraries. Journal
of Digital Information, 6(3). Retrieved October 9, 2006, from
http://jodi.tamu.edu/Articles/vO6/iO3/Fast/

Fentress, J ., & Wickham, C. (1992). Social memory. Oxford UK: Blackwell.

Fertig, S., Freeman, E., & Gelemter, D. (1996). “Finding and reminding” reconsidered.
SIGCHI Bulletin, 28(1), 66-69.

F leckenstein, K. S. (1999). Writing bodies: Somatic mind in composition studies.
College English, 61(3), 281—306.

Flower, L. (1989). Cognition, context, and theory building. College Composition and
Communication, 40(3), 282-311.

Flower, L., & Hayes, J. R. (1981). A cognitive process theory of writing. College
Composition and Communication, 32(4), 365-387.

F rancoz, M. J. (1999). Habit as memory incarnate. College English, 62(1), 11-29.

Fussell, S. R., Kraut, R. E., & Siegel, J. (2000). Coordination of communication: Effects
of shared visual context on collaborative work. Proceedings of CSC W 2000,
December 2-6, 2000, Philadelphia, PA, 21-30.

Gee, J. P., Hull, G., & Lankshear, C. (1996). The new work order: Behind the language
of the new capitalism. Sydney: Westview Press.

Glosiene, A., & Manzhukh, Z. (2005). Towards a usability framework for memory
institutions. New Library World, 106(1214/1215), 303-319.

Goodwin, C., & Goodwin, M. H. (1996). Seeing as a situated activity: Formulating
planes. In Engestrom Y., & Middleton D. (Eds.), Cognition and communication
at work (pp. 61-95). New York: Cambridge UP.

Haas, C. (1996). Writing technology: Studies on the materiality of literacy. Mahwah,
NJ: Lawrence Erlbaum Associates.

169

 

Haas, C., & Witte, S. P. (2001). Writing as an embodied practice: The case of
engineering standards. Journal of Business and Technical Communication,
15(4), 413-457.

Hart-Davidson, W. (2005). Shaping texts that transform: Toward a rhetoric of objects,
relationships, and views. In M. Day & C. Lipson (Eds.), Technical
communication and the World Wide Web (pp. 27-42). Mahwah, NJ: Lawrence
Erlbaum Associates.

Hayes, J. R. (1996). A new framework for understanding cognition and affect in
writing. In C. M Levy & S. Ransdell (Eds.), The science of writing: Theories,

methods, individual diflerences, and applications (pp. 1-27). Mahwah, NJ:
Lawrence Erlbaum Associates.

Hayne, S. C. (2005). The use of pattern-communication tools and team pattern
recognition. IEEE Transactions on Professional Communication, 48(4), 377-
390.

Hays, P. A. (2004). Case study research. In K. deMarrais & S. D. Lapan (Eds.),
Foundations for research: Methods of inquiry in education and the social
sciences (pp. 217-234). Mahwah, NJ: Lawrence Erlbaum Associates.

Hedstrom, M. (2002). Archives, memory, and interfaces with the past.
Archival Science, 2, 21-43.

Hobbs, C. L. (2002). Rhetoric on the margins of modernity: Vico, Condillac,
Monboddo. Carbondale: Southern Illinois UP.

Hong, W., Thong, J. Y. L., Wong, W., & Tarn, K. (2002). Determinants of user
acceptance of digital libraries: An empirical examination of individual
differences and system characteristics. Journal of Management Information
Systems, 18(3), 97-124.

Hutchins, E. (19953). Cognition in the wild. Cambridge, MA: MIT Press.

Hutchins, E. (1995b). How a cockpit remembers its speeds. Cognitive Science, 19, 265-
288.

Johnson, R. R. (1998). User-centered technology: A rhetorical theory for computers
and other mundane artifacts. Albany: SUNY Press.

J ohnson-Eilola, J. (2005). Datacloud: Toward a new theory of online work. Hampton
Press.

170

 

Kani-Zabihi, E., Ghinea, G, & Chen, S. Y. (2006). Digital libraries: What do users
want? Online Information Review, 30(4), 395-411.

Kaptelinin, V., & Nardi, B. A. (2006). Acting with technology: Activity theory and
interaction design. Cambridge, MA: MIT Press.

Kellogg, R. T. (1996). A model of working memory in writing. In C. M. Levy & S.
Ransdell (Eds.), The science of writing: Theories, methods, individual
diflerences, and applications (pp. 57-71). Mahwah, NJ: Lawrence Erlbaum
Associates.

Kinneavy, J. L. (1986). Kairos: A neglected concept in classical rhetoric. In J. D. Moss
(Ed.), Rhetoric and praxis: The contribution of classical rhetoric to practical
reasoning (pp. 79-105). Washington, DC: The Catholic University of America

Press.

Kinneavy, J. L. (2002). Kairos in classical and modern rhetorical theory. In P.
Sipiora & J. S. Baumlin (Eds.), Rhetoric and kairos: Essays in history, theory,
and praxis (pp. 58-76) Albany: SUNY Press.

Kirsh, D. (1995). The intelligent use of space. Artificial Intelligence, 73, 31-68.

Krarnp, M. K. (2004). Exploring life and experience drrough narrative inquiry. In K.
deMarrais & S. D. Lapan (Eds.), Foundations for research: Methods of inquiry
in education and the social sciences (pp. 103-121). Mahwah, NJ: Lawrence
Erlbaum Associates.

Krishnan, A., & Jones, S. (2005). TimeSpace: Activity-based temporal visualization of
personal information spaces. Personal and Ubiquitous Computing, 9, 46-65.

Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied mind and its
challenge to western thought. New York: Basic Books.

Lansdale, M. W., & Oliff, L. (2005). Quantifying precision and availability of location
memory in everyday pictures and some implications for picture database design.
Journal of Experimental Psychology: Applied, 1 1(2), 67-83.

Larsen, R. L., & Wactlar, H. D. (2003). Knowledge lost in information: Report of the
NSF workshop on research directions for digital libraries. NSF Post Digital
Library Futures Workshop, Chatham, Mass, June 15-17, 2003. Retrieved
August 2, 2007, from http://www.sis.pitt.edu/~dlwkshop/report.pdf

Latour, B. (1987). Science in action: How to follow scientists and engineers through
society. Milton Keynes: Open University Press.

171

 

Lauer, J ., & Asher, J. W. (1988). Composition research: Empirical designs. New
York: Oxford UP.

Leander, K., & Prior, P. (2004). Speaking and writing: How talk and text interact in
situated practice. In C. Bazerman & P. Prior (Eds.), What writing does and how
it does it: An introduction to analyzing texts and textual practices (pp. 201 -237).
Mahwah, NJ: Lawrence Erlbaum Associates.

LeFevre, K. B. (1987). Invention as a social act. Carbondale: Southern Illinois UP.

Luria, A. R. (1987). The mind of a mnemonist. A little book about a vast memory. (L.
Solotaroff, Trans.) Cambridge, MA: Cambridge UP.

Lyotard, J. (1999). The Postmodern Condition: A Report on Knowledge. (G.
Bennington & B. Massumi, Trans). Minneapolis: University of Minnesota
Press. (Original work published in 1979)

Malone, T. W. (1983). How do people organize their desks? Implications for the design
of office information systems. ACM Transactions on Ofi‘ice Information Systems
1(1), 99-112.

Matusiak, K. K. (2006). Towards user-centered indexing in digital image collections.
OCLC Systems & Services, 22(4), 283-298.

McNeil], D. (1992). Hand and mind: What gestures reveal about thought. Chicago:
University of Chicago UP.

McNeill, D. (2005). Gesture and thought. Chicago: University of Chicago UP.

Miles, M. B., & Huberrnan, A. M. (1994). Qualitative data analysis: An expanded
sourcebook (2nd ed.). Thousand Oaks, CA: Sage Publications.

Murphy, J. J. (2001). The key role of habit in Roman writing instruction. In J. J.
Murphy (Ed.), A short history of writing instruction (2"d ed.) (pp. 35-78).
Mahwah, NJ: Lawrence Erlbaum Associates.

Murphy, J. J. (2002). The metarhetoric of Aristotle, with some examples from his 0n
Memory and Recollection. Rhetoric Review, 21(3), 213-228.

Neuwirth, C. M., & Kaufer, D. S. (1989). The role of external representations in the
writing process: Implications for the design of hypertext-based writing tools.

Hypertext ‘89 Proceedings, 319-341.

Norman, D. A. (2002). The design of everyday things. New York: Basic Books.

172

 

Ong, W. J. (1982). Orality & literacy: The technologizing of the word. New York:
Routledge.

Perry, M. J ., Fruchter, R., & Rosenberg, D. (1999). Co-ordinating distributed
knowledge: A study into the use of an organizational memory. Cognition,
Technology, & Work, 1, 142-152.

Plato. (1995). Phaedrus. (A. Neharnas & P. Woodruff, Trans). Indianapolis, IN:
Hackett Publishing.

Porter, J. E., Sullivan, P., Blythe, S., Grabill, J. T., & Miles, L. (2000). Institutional
critique: A rhetorical methodology for change. CCC 51.4.

Prior, P. (2004). Tracing process: How texts come into being. In C. Bazerrnan & P.
Prior (Eds.), What writing does and how it does it: An introduction to analyzing
texts and textual practices (pp. 167-200). Mahwah, NJ: Lawrence Erlbaum
Associates.

Prior, P., & Shipka, J. (2003). Chronotopic lamination: Tracing the contours of literate
activity. In Bazerrnan & Russel (Eds.), Writing selves/writing
societies: Research from activity perspectives. Fort Collins, Colorado: The
WAC clearinghouse and mind, culture, and activity. Retrieved March 24, 2006,
from http://wac.colostate.edu/books/selves_societies/

Reich, R. B. (1991). The work of nations: Preparing ourselves for 21st century
capitalism. New York: Vintage Books.

Reynolds, R. E., Sinatra, G. M., & Jetton, T. L. (1996). Views of knowledge acquisition
and representation: A continuum from experience centered to mind centered.
Educational Psychologist, 31(2), 93-104.

Schwaber, K., & Beedle, M. (2002). Agile software development with scrum. Upper
Saddle River, NJ: Prentice Hall.

Selfe, C. L., & Selfe, R. J. Jr. (1994). The politics of the interface: Power and its
exercise in electronic contact zones. College Composition and Communication,
45(4): 480-504.

Sellen, A. J ., & Harper, H. R. (2002). The myth of the paperless office. Cambridge, MA:
MIT Press.

Sharples, M. (1996). An account of writing as creative design. In C. M. Levy & S.
Ransdell (Eds.), The science of writing: Theories, methods, individual
differences, and applications (pp. 127-148). Mahwah, NJ: Lawrence Erlbaum

Associates.

173

 

Sheard, C. M. (1993). Kairos and Kenneth Burke’s psychology of political and social
communication. College English, 55(3): 291-310.

Slack, J. D., & Wise, J. M. (2005). Culture + technology: A primer. New York: Peter
Lang.

Small, J. P. (1997). Wax tablets of the mind: Cognitive studies of memory and literacy
in classical antiquity. New York: Routledge.

Smith, J. E. (2002). Time and qualitative time. In P. Siporia & J. S. Baurnlin (Eds.),
Rhetoric and kairos: Essays in history, theory, and praxis (pp. 46-57). Albany,
NY: SUNY Press.

Sorabji, R. (1972). Aristotle on memory. Providence: Brown UP.

Spinelli, G., Perry, M., & O’Hara, K. (2005). Understanding complex cognitive
systems: The role of space in the organization of collaborative work. Cognition,
Technology, and Work, 7, 111-118.

Spinuzzi, C. (2001). Sofiware development as mediated activity: Three analytical
frameworks for studying compound mediation. Proceedings of A CM SIGDOC .
New York, NY: ACM Press, 58-67.

Spinuzzi, C. (2003a). Compound mediation in software development: Using genre
ecologies to study textual artifacts. In C. Bazennan & D. Russell, Writing
selves/writing societies: Research from activity perspectives. Fort Collins,
Colorado: The WAC Clearinghouse. Retrieved January 10, 2008, from
http://wac.colostate.edu/books/selves_societies/

Spinuzzi, C. (2003b). Tracing genres through organizations: A sociocultural approach
to information design. Cambridge, MA: MIT Press.

Star, S. L. (1999). The ethnography of infrastructure. American Behavioral Scientist,
43(3), 377-391.

Steel, C. (2006). Roman oratory. Cambridge: Cambridge University Press.

Swarts, J. (2004). Cooperative writing: Achieving coordination together and
apart. Proceedings of the 2004 ACM Conference on Documentation, 83-89.

Thayer, S. M., & Steenkiste, P. (2003). An architecture for the integration of physical
and informational spaces. Personal and Ubiquitous Computing, 7, 82-90.

Wenger, E., McDerrnott, R., & Snyder, W. M. (2002). Cultivating communities of

practice: A guide to managing knowledge. Boston: Harvard Business School
Press.

174

 

Winsor, D. A. (2000). Ordering work: Blue-collar literacy and the political nature of
genre. Written Communication, 17(2), 155-184.

Winsor, D. A. (2003). Writing power: Communication in an engineering center.
Albany: State University of New York Press.

Yates, J. (1989). Control through communication: The rise of system in American
management. Baltimore: Johns Hopkins UP.

Yin, R. K. (2003). Case study research: Design and methods (3rd ed). Thousand
Oaks, CA: Sage Publications.

Zhang, J ., & Marchionini, G. (2005). Evaluation and evolution of a browse and search
interface: Relation browser-H: Proceedings of the 2005 National Conference on
Digital Government Research, Atlanta, GA, 179-188.

Zuboff, S. (1988). In the age of the smart machine: The future of work and power.
New York: Basic Books.

175