-
A Semiotic Analysis of Corporate Language: Organizational
Boundaries and Joint VenturingAuthor(s): C. Marlene FiolSource:
Administrative Science Quarterly, Vol. 34, No. 2 (Jun., 1989), pp.
277-303Published by: Sage Publications, Inc. on behalf of the
Johnson Graduate School of Management, CornellUniversityStable URL:
http://www.jstor.org/stable/2989899 .Accessed: 18/06/2013 16:19
Your use of the JSTOR archive indicates your acceptance of the
Terms & Conditions of Use, available at
.http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars,
researchers, and students discover, use, and build upon a wide
range ofcontent in a trusted digital archive. We use information
technology and tools to increase productivity and facilitate new
formsof scholarship. For more information about JSTOR, please
contact [email protected].
.
Sage Publications, Inc. and Johnson Graduate School of
Management, Cornell University are collaboratingwith JSTOR to
digitize, preserve and extend access to Administrative Science
Quarterly.
http://www.jstor.org
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
A Semiotic Analysis of Corporate Language: Organizational
Boundaries and Joint Venturing
C. Marlene Fiol New York University
? 1989 by Cornell University. 0001 -8392/89/3402-0277/$1
.00.
0
Many people deserve recognition for helping me present the
narrative semiotic method used in this study. The credit for
achieving the present level of clarity is largely theirs: Stephen
Barley, Janice Beyer, Janet Dukerich, Roger Dunbar, Jane Dutton,
William Guth, James Hoover, Anne Huff, Marjorie Lyles, Frances Mil-
liken, Linda Pike, and the editor and anon- ymous reviewers of
Administrative Science Quarterly. This research was par- tially
supported by a grant from the Ten- neco Fund.
In an attempt to explain differences in the propensity of
organizations to enter into joint ventures, this study uses a
semiotic method of textual analysis to examine CEOs' letters to
shareholders as they reflect the existence and strength of
boundaries separating internal organizational subunits and
boundaries separating the company from external environments. These
boundaries are linked to the propensity of ten companies in the
chemical industry to engage in joint-venture activity. Weak
internal or interdi- visional boundaries and strong boundaries
separating the company from the external environment were prevalent
in firms having no joint ventures; strong interdivisional
boundaries and weak external boundaries were prevalent in firms
having several joint ventures."
ORGANIZATIONAL BOUNDARIES AND JOINT VENTURES Organizational
boundaries are imagined lines drawn to sepa- rate the organization
from its surrounding environment and to specify how internal roles
and functions are related but also separated from one another
(Wilden, 1980). Corporate boundaries thus define and delimit
corporate relationships, delineating areas of autonomy and
self-control as well as areas of interdependence. Theories of group
dynamics (Miller and Rice, 1967; Miller, 1979) suggest that
perceptions of boundaries and autonomy can be related circularly:
as au- tonomy is threatened, boundaries that separate a unit from
its surroundings are strengthened, which leads to increased au-
tonomy and perceptions of greater self-control. Organizational
boundary definitions have implications for issues of control,
particularly with respect to joint-venture strategies.
Joint-venture relationships require shared control of particular
activities by two or more companies. Berg, Duncan, and Friedman
(1982) argued that many firms do not enter joint ventures because
their CEOs fear loss of control. Harrigan (1985) suggested that
organizations may fear that sharing control is an admission of
weakness. Consistent with earlier research on joint ventures, this
study argues that CEOs who prefer autonomy (control) will be less
likely to enter into joint ventures. One would expect them to view
their firms as separate from the environment and to draw strong
external boundaries around their firm. Firms that enter
partnerships with others, however, would be expected to tol- erate
a more open exchange between their organization and the environment
and to encourage weak external boundaries. Joint ventures also
require tolerance for organizational sepa- rateness of internal
units, so that the new-venture "child" can be distinguished from
the "parents" (Harrigan, 1985). This implies relatively strong
internal boundaries that differentiate the joint-venture units from
other units within each of the partners, leading to greater
internal separateness. Firms without joint ventures would be
expected to encourage rela- tively cohesive internal units, with
weak internal boundaries. In contrast, joint-venturing firms would
tolerate relatively au- tonomous organizational subunits, drawing
strong internal boundaries. Finally, if tendencies to draw or
remove strong boundaries between one's own and other organizations
are manifesta- 277/Administrative Science Quarterly, 34 (1989):
277-303
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
tions of a more general disposition, one would expect them to be
reflected in the way an organization communicates facts about
itself in relation to its environment. Joint-venturing firms would
use language conveying external inclusiveness and shared fate and
relatively loose control of internal sub- units. Firms discouraging
joint ventures would speak a lan- guage of external separation and
disassociation and internal inclusiveness. This study's central
proposition is as follows: Proposition: Joint-venturing firms are
likely to exhibit weak bound- aries separating the company from
external environments and strong boundaries among organizational
subunits. Conversely, non- joint-venturing firms are likely to
exhibit strong boundaries separating the company from external
environments and weak boundaries among organizational subunits.
While letters to shareholders directly communicate facts about a
firm, they also communicate implicit beliefs about the organization
and its relationships with the surrounding world. A semiotic method
of textual analysis is used in this paper to analyze the narrative
structures of letters to shareholders and identify consistent
patterns in these implicit beliefs. The method uncovers systematic
differences in the way these documents portray the boundaries that
define interrelation- ships among units within the company and
between the company and external environments. These differences
are then shown to distinguish between firms involved in do- mestic
joint ventures and those that are not. THE SEMIOTICS OF
ORGANIZATIONAL BOUNDARIES A firm's posture toward joint-venture
activities can be seen as part of a more general pattern of beliefs
regarding the appro- priate way to define internal and external
relationships. Pat- terns of beliefs are grounded in underlying
meanings attached to self and others (Greimas, 1987). Identifying
such patterns requires a methodology that can detect meanings
assigned to events and situations and specify the rules that govern
meaning in a given context. Semiotics is one such method- ology.
Semiotics is a formal mode of analysis used to identify the rules
that govern how signs convey meanings in a particular social system
(Eco, 1979). With roots in the structural lin- guistic principles
of Saussure (1916), it uses linguistic sign systems as a model to
identify and make explicit the rules governing other systems of
signification, e.g., symbolic be- haviors or physical trappings.
The signs that combine to convey meaning in natural language
systems are words. Grammatical rules govern their meaningful
combinations. Se- miotics assumes that a similar set of rules or
social conven- tions governs the choices and constraints that
determine expressions of meaning in other social systems. For
example, manner of dress is a sign system conveying social
meanings. A single sign, such as a woman's hat, may variously
denote modesty, religiosity, or style, depending on the conventions
of a particular context. Different signs within the same system
(other articles of clothing) may denote the same meaning, again
based on social conventions. Semiotics assumes that diverse signs
or expressions can convey shared meaning because they are grounded
in a common set of underlying values. It further assumes that a
278/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
value can be defined only in relation to its logical opposites.
A frequently used illustration (Greimas, 1966) is that "love" has
no meaning without the opposed concepts of "hate," "non- love," and
"non-hate." Thus the rules that regulate meaning are based on
underlying oppositions that define the values embedded in the
system. Different forms of semiotics have emerged for the study of
different kinds of sign systems. Though all forms of the method
assume that the meaning of individual signs is grounded in
underlying value oppositions, they differ in the manifestation of
signs that is the focus of study. Saussurean semiotics (Saussure,
1916; Eco, 1979) examines the links be- tween natural language
signs and the meanings they signify. A derivative of Saussure's
model has been applied to the analysis of behavioral signs and
their meanings. By semioti- cally analyzing the rules that regulate
the meaning of work in a funeral home business, for example, Barley
(1983) demon- strated that a recurrent underlying value-the denial
of death -permeated the firm's culture and determined how all work
decisions and actions took place. Narrative, or Greimassian
semiotics (Greimas, 1966), another form of the method, treats a
narrative or group of narratives as the sign system. Its aim is to
uncover how narrative com- ponents combine to convey meanings
within texts. Narrative semiotic analyses of myths, folktales, and
other stories re- lating to social life have been used extensively
by anthropolo- gists and sociologists in studies of cultures and
interpretive systems (Propp, 1958; Levi-Strauss, 1963, 1976;
Manning, 1987). Levi-Strauss, for example, found that shared
meanings of multiple and variable myths can be uncovered by identi-
fying stable patterns in the way the stories' components are
structured. This study uses narrative semiotic techniques to
uncover pat- terns of internal and external boundary definitions in
letters to shareholders. The letters are a particularly useful data
set for semiotically examining beliefs about boundaries, because
(1) they share the common purpose of conveying a positive corporate
image to an important constituency, (2) the CEO approves the image
that is conveyed, and (3) their production is ruled by a fairly
rigid set of conventions governing what must be addressed and in
what form. These corporate mes- sages are like folktales in that
they tell many stories that can be reordered to extract a set of
recurring structures that re- flect underlying values.
Narrative Semiotic Method Classic narrative theory, as set forth
by Russian Formalists (Propp, 1958), defines a narrative as a set
of interlocking signs whose meanings are determined by underlying
rules that reg- ulate how different units of texts may be combined.
A narra- tive consists of the following elements: a narrative
subject in search of an object, a destinator (an extratextual
force, the source of the subject's ideology), and a set of forces
that ei- ther help or hinder the subject in acquiring the desired
object. These elements are not actors in the traditional sense. For
example, forces of help may consist of the subject's frame of mind,
the activities of the subject, or the actions of others in the
story. 279/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
I The italicized terms are imperfect transla- tions of Greimas's
Pouvoir and Savoir (Greimas, 1966).
The stories told in letters to shareholders narrate the suc-
cesses and failures of individuals (e.g., the CEOs), organiza-
tional subunits, and the entire company; the narrative subject may
be any one of these. The ultimate object of the narrative is in all
cases at the firm level and involves the projection of a positive
company image to shareholders. The destinator is the corporate and
societal environment of the late 1970s, the context within which
the subjects of these narratives are operating. The forces of help
and hindrance are variably de- picted as any of a larger set of
potential internal and external narrative forces. The narrative
progresses toward a resolution through the re- lations developed
between the subject and other narrative forces. These relations are
distinguished on the basis of what is called in semiotics their
tonality and modality, narrative components qualifying and
characterizing the relations among narrative forces. The tonality
of a narrative represents the judgment of the subject on a
phenomenon and can be posi- tive or negative. The modality of a
narrative distinguishes be- tween the doing and being of the
subject and may be pragmatic or cognitive. A pragmatic mode
establishes the subject's being-able-to-do, or ability to act.
Pragmatic rela- tions between subject and other narrative forces
take place within action sequences of the story. A cognitive mode
es- tablishes the subject's knowing-how-to-do, or state of mind.
Cognitive relations take place outside of action sequences.' Figure
1 depicts the forces and relational elements assumed to exist in a
narrative. The modalities are being-able-to-do and
knowing-how-to-do. The narrative forces are in boxes. The re-
lationships between them are depicted in the connecting lines.
Cognitive relations develop through the subject's in- terior
discourse and represent one source of help and hin- drance in
achieving the object. Pragmatic relations develop through the
subject's action and represent the other source of help and
hindrance. Positive or negative tone denotes whether the narrative
force with which the subject is inter- acting is a help or a
hindrance. Ultimately, to uncover the structural patterns of a
narrative, one must identify and classify four components of
narrative structure (Greimas, 1966). The first two components
define what relations exist between the subject and other narrative
forces, and the second two define how those relations func- tion in
the narrative plot: (1) the degree of association among the forces
that can serve as narrative subjects, (2) the spatial association
between all internal and external forces involved in the subject's
quest, (3) the mode of the subject and spatial associations as
either pragmatic or cognitive, and (4) the tone of the subject and
spatial associations as either positive or negative. The procedure
for identifying these components in the letters to shareholders was
a three-phase narrative anal- ysis that proceeded as follows: The
aim of Phase 1 was to identify the apparent story line and goals of
the letters, the surface structure of the text. The unit of
analysis was the letter in the exact sequence and form it was
published. Phase 2 involved three steps to derive the narrative
struc- ture of the text:
280/ASQ, Ju ne 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries Figure 1. Narrative forces and
relations.
Destinator
Cognitive Help Hindrance Relations
Knowing-How-to-Do
Subject
Positive Negative Relations Relations
Pragmatic Help Hindrance Relations Being-Able-to-Do
Objct
Step 1. Topic blocks (called programs) were identified in the
letters. A program is a block of narrative that is classified as
having a single narrative goal, e.g., to explain capital expendi-
tures or to explain losses. Programs represented the primary unit
of analysis of this phase of the study. Step 2. Each program was
characterized on the basis of the four narrative-structure
components discussed above. Each of these components was classified
within an oppositional framework (e.g., the narrative subject may
have been pre- sented as a single actor (CEO) or an aggregated set
of actors (the entire industry); the tone may have been negative or
positive). Step 3. The interactions of the four components were
traced to show how they combined to achieve the program's goal
(e.g., negative tone combined with a single-actor subject in
describing the company's losses). The combined set of inter-
actions in each program is called the program's oppositional
structure. Phase 3 involved further combining the oppositions to
derive an oppositional map for an entire letter, or the letter's
deep structure. The deep structure consists of a dominant under-
lying value in the text and the value's logical opposites, e.g.,
the value of strong external boundaries and its opposites.
Narrative rules govern the links between the intermediary narrative
component structures found in Phase 2 and the value schema of Phase
3. The consistency with which the rules are applied reflects the
extent to which underlying meanings are patterned in these letters.
The analysis moves from a focus on the texts as a whole, to their
smallest narrative units (the narrative forces), to succes- sive
combinations of these forces until the focus is again on the entire
texts. This movement represents the core of the process of
determining the meaning of a text: deconstruction of meanings as
they appear in isolation and subsequent re- construction of those
meanings as they relate to others in the narrative system. 281/ASQ,
June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
An overview of the three analytic phases is provided in Table 1,
which outlines the major assumptions, the unit and the aim of
analysis, and the procedure for deriving the three levels of
textual structure. A fuller description of the method is pro- vided
below as it applies to this study's sample of texts.
Table 1
Overview of the Method
Phase and level of Major Unit of Aim of structure assumption
analysis analysis Procedures
1. Surface Intended goal Entire letter as Identify intended
Read, compare, congruity within series of words. goals of letter.
and code and among intended letters. narrative goals of
texts in sample. 2. Narrative Multiple subgoals Step 1:
Programs. Identify programs, Break each letter
within a single Step 2: each with a into programs letter.
Oppositional subgoal. with narrative
Narrative forces narrative Identify subgoals. combine components
oppositional Trace positions of differentially to (subjects,
narrative oppositional achieve environments, components for
narrative subgoals. mode, tone). each program. components in
Step 3: Combined Identify combined each program. configuration
of narrative Cross-classify multiple narrative structure for
positions of components. each program. narrative
components to arrive at combined structure.
3. Deep Structural Underlying set of Identify combined Derive
combined oppositions oppositions for oppositional structure
underlying entire entire letter. structure for through letter
traceable entire letter. combinations in through Step 3.
configuration of multiple narrative structures.
Before applying a narrative analysis, the researcher must
identify an appropriate deep-structure schema using two cri- teria:
(1) The schema must be meaningful in light of the study's purpose,
and (2) it must encompass the recurring op- positional structures
delineated in Phase 2 of the process. This preliminary
deep-structure search is a form of pretest, using texts similar to
but not including the actual study sample. The researcher carries
out an iterative search, in- volving all three analytic phases,
until an appropriate, suffi- ciently limited, yet encompassing set
of oppositions is found. Once a schema has been identified, Phases
1 and 2 are ap- plied to each text in the chosen sample of
narratives to de- termine the position and movements of the
narrative structures along the oppositional axes of the
deep-structure schema that is assumed for that group of texts. In
this study, I used an iterative search process, applying Phases 1
through 3 to a number of texts that were similar to
282/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
the sample of texts eventually used in that they were letters to
shareholders of other chemical companies' annual reports. The
depiction of internal and external organizational bound- aries in
those letters signified oppositions that met the two criteria for
appropriateness: They were meaningful for the study's purpose of
examining boundaries (first criterion), and they could be organized
into an oppositional deep-structure schema that encompassed the
recurring narrative structures in the texts (second criterion). The
semiotic square, depicting this oppositional deep structure, is
presented in Figure 2.
Figure 2. Semiotic square of oppositional deep-structure schema.
(S) (Si) Weak Weak External Internal Boundaries 3 Boundaries
1 5 2
Strong Strong Internal 4 External Boundaries Boundaries (-Si)
(-S)
'Numbers on the axes refer to internal and external boundary
combi- nations: 1 = Weak external/Strong internal boundaries 2 =
Weak internal/Strong external boundaries 3 = Weak external/Weak
internal boundaries 4 = Strong external/Strong internal boundaries
5 = Strong internal/Weak internal and/or Weak external/Strong
ex-
ternal boundaries
The dominant positive value of the system being analyzed is
positioned in the upper-left corner of the square. Logical rela-
tions of implication, contrariety, and contradiction govern the
positions of the other three values. In this case, relations be-
tween internal and external are relations of contrariety; rela-
tions between weak and strong are relations of contradiction. The
negative of a value (its contradiction) is not identical to its
contrary. For example, the square in Figure 2 assumes that the text
emphasizes weak external boundaries (S) as the pri- mary positive
value. Its negation or denial (- S) implies but is not identical to
its contrary (Si); similarly, the negation of the negative value (-
Si) implies but is not equivalent to the pos- itive value (S). The
axes of a semiotic square are characterized by the two values they
encompass. Each narrative sequence (program) of the letters to
shareholders embraces one of these value combinations and may be
located on an axis of the square. Thus, a program's position in the
square describes the under- lying values in that piece of the text.
Moreover, it indicates a set of rules (discussed below) that
dictate how the four nar- rative components may combine in each
program to develop the story with that particular set of underlying
values. 283/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
If external weak boundaries represent the dominant positive
value of a text (assumed in the square in Figure 2), programs
reflecting Position 1 tell their story through depicting exter-
nally weak and the logically implied internally strong bound-
aries. Position 2 reflects contradictions of both values of
Position 1. Position 3 embraces what semiotics calls a com- plex
set of contrary values, since they contradict the internal boundary
definition implied by the dominant value in the text. Position 4
embraces what semiotics calls neutral contraries, since they
contradict Position 1 only by the inclusion of the external
boundary definition implied in the nondominant Posi- tion 2.
Position 5 is an inherently contradictory state, indi- cating
structures that destroy themselves. Though a narrative may (and
often does) contain contradictory structures, the conflict in the
narrative is untenable if the narrative consis- tently remains in
this position (e.g., the narrative conflict for the hero of
folktales cannot be established if the entire story simultaneously
embraces values of love and non-love or hate and non-hate). The
semiotic square may also be used dynamically, in that the logical,
atemporal relations of value can generate a temporal sequence
(Haidu, 1982). An entire narrative may remain con- sistent with a
single set of values, indicating no change; it may begin with one
set and end with another, indicating change; or it may be circular,
where transformations lead the final state back to the beginning,
indicating the loss of a po- tential for change (Haidu, 1982). With
the internal/external boundary schema as a given deep- structure
model for this study's texts, the subsequent anal- ysis involved
locating each narrative program on one of the axes of the square.
The configuration of these positions in the square illustrates the
dynamics of an underlying pattern of boundary definitions in the
texts. It then becomes possible to differentiate texts whose
structures reflect very different values of internal and external
boundaries even while de- scribing the pursuit of similar
goals.
SEMIOTIC ANALYSIS OF LETTERS TO SHAREHOLDERS A narrative
semiotic analysis was performed on CEOs' letters to shareholders in
ten chemical companies' annual reports of 1977-1979. Five companies
had two or more domestic joint ventures and five had none. Ten
firms were chosen that have similar product lines, markets, and
sizes. All were heavily in- volved in agricultural and industrial
markets and were begin- ning to shift toward high technology and
specialty chemical markets. Mann-Whitney U Tests indicate that the
joint- venturing and non-joint-venturing firms were not
significantly different (at = .10) in terms of level of sales,
relative capital expenditures, or relative net income. However, the
ten firms were significantly different in terms of relative
R&D, joint- venturing firms spending relatively less on
internal R&D. This supports economists' (Berg, Duncan, and
Friedman, 1982) claims that domestic joint ventures in the chemical
industry serve as substitutes for internal R&D. Summary
economic profiles of the ten companies are presented in Table 2.
Because semiotic coding relies on the researcher's assess- ments in
applying the structural framework and coding the
284/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
texts, tests of coding reliability are necessary. For this
study, an independent researcher, trained in semiotics by the
author but unaware of the study's propositions or the author's re-
sults, replicated the entire analysis. Of the 150 programs (3
years, 5 programs in each year's letter, for 10 companies), the
semiotic positioning of only three differed from the author's own
analysis. In the final compilation of the deep-structure results,
no results differed from those presented here.
Table 2
Economic Profile of Companies in Sample*
Sales Capital Company (Millions) R&D/Sales
Expenditures/Sales Profit/Sales Two or more domestic
joint ventures A 1489 .015 .092 .039 B 1279 .005 .049 .07 C 4191
.008 .082 .04 D 1781 .02 .089 .053 E 1235 .004 .13 .085
Non-joint- venturing
V 1211 .025 .148 .089 W 613 .008 .034 .024 X 2412 .03 .099 .042
Y 816 .025 .054 .043 z 1158 .037 .095 .055
* Figures are all averages of ten years (1972-1981).
Before beginning Phase 1 of the analysis, interviews were
conducted with a senior-level executive in charge of new ventures
(with tenure of at least ten years) in each of the firms. The
interviews served three purposes: (1) They con- firmed the accuracy
of published reports about the joint- venture activity or
inactivity of the firms; (2) they verified that no company in the
sample had experienced a change in CEO or a major restructuring
during a ten-year period preceding 1977, the first year of the
study; and (3) they established the environmental context within
which the letters were written. There was marked consensus among
all informants on the perceived need in the late 1970s to show
shareholders that the companies were shifting their focus from the
mature in- dustrial chemicals business to high-growth specialty
chem- icals.
Surface Structure
The surface structure of the letters consists of the words and
sentences as these have been written and arranged by the authors.
The letters reflect the conscious aim of the authors, as stated in
the interviews, to present a positive image of the company to its
stockholders. The surface structure of the two letters to be
analyzed in detail, Company E (joint-ven- turing) and Company Z
(non-joint-venturing) is presented in its entirety in the Appendix,
with the program codings, or topic blocks, that were identified for
the semiotic analysis.
285/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Narrative Structure The researcher carrying out a narrative
structural analysis pro- ceeds in three steps (Greimas, 1966): (1)
identify a set of topic blocks or programs within the letters; (2)
identify the narrative components of each topic block; and (3)
delineate the interactions of the narrative components that will
point to the underlying structural configurations that are the
narrative structures. This procedure was followed in this study.
Step 1. The letters were first divided into programs, each one
directed toward a single implied narrative subgoal that the subject
wishes to attain. The goals of the programs in the texts of these
companies over the three years follow a similar pattern. The
letters usually begin with a program that serves as an intertextual
bridge between the previous year's letter and the current one.
Usually this entails outlining the com- pany's strengths and
special growth areas (designated as Program 1). Its function in the
story is to outline the subject's current knowledge and competence
to achieve the final ob- ject, a positive company image. Program I
is most often fol- lowed by a program giving reasons for any
weaknesses or poor performance (Program 11). Its function is to
establish the conflict of the story by outlining the knowledge or
compe- tence that may still be lacking. The third program discusses
expenditures on capital assets and R&D (Program 111), and the
fourth describes changes in management (Program IV). Their function
in the story is to show the development of assets that will resolve
the conflicts introduced in Program I1. Finally, the future outlook
of Program V functions as the story's de- nouement. The programs of
the 1977 letters of illustrative companies E and Z are blocked out
in the Appendix. Step 2. In this step, the narrative components
were identified that distinguish the programs of any narrative text
(subject relations, subject/environment relations, mode, and tone)
in a semiotic analysis, and their oppositional positions were lo-
cated within the narrative. The results of the narrative-structure
phase of the analysis are presented in Table 3. Of the ten
companies of the sample, companies A-E are joint-venture active,
companies V-Z are not joint-venture active. Companies Y and Z are
listed twice because the structure of their letters changed between
1977 and 1979. Each company's letters are described in two rows.
The "S" row describes relations among possible subjects; the "E"
row describes subject/environment relations. The table lists the
five programs of each letter and describes the narrative components
identified for each program. The fol- lowing illustration of the
coding procedure applies to compa- nies E and Z-77, which are shown
in darker type in Table 3. Subject relations. The subjects of these
narratives may be depicted in varying degrees of aggregation,
ranging from the CEO or any other individual member of the company
to the company as a whole. The extent of unity portrayed among
various possible subjects indicates the extent of inclusion of
multiple levels of aggregated forces into a single narrative
subject. Subject relations in these stories were coded on the basis
of the use of nouns and pronouns to depict the narrative sub-
286/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
Table 3
Oppositional Narrative Forces
Program Company I 11 Ill IV V Joint-venturing A S D-pNT AD-cNT
AD-pNT AD-pP D-pNT
E AD-pP A-cNT A-pP A-pP B S D-pNT D-pNT D-pNT AD-pNT
E A-pP A-pP A-pP AD-pNT C S D-pNT D-pNT D-pNT D-pP D-pNT
E AD-pNT AD-pNT AD-pNT D S D-pNT D-pcNT AD-pNT D-pNT D-pNT
E AD-pP A-pcP AD-pP E S D-pcNT AD-pcP D-pcNT D-cNT
E A-pcP A-pcN Non-joint-venturing Y-79 S D-cNT D-cNT D-pcN D-cNT
D-cNT
E A-cP A-pcNT A-cNT Z-79 S D-pNT D-pNT D-pNT D-pNT
E AD-pP AD-pP AD-pP D-pP V S A-pP D-pcNT AD-pNT A-pP AD-pcNT
E D-pNT D-pcN D-pNT D-pcNT W S AD-pP D-cNT A-pcP AD-pNT
A-pNT
E D-cNT D-pcNT D-pN X S AD-pP AD-cN AD-pP AD-pP A-pNT
E AD-pNT D-cNT D-pNT Y-77 S A-cP A-cNT A-cNT
E D-cN Z-77 S A-pP A-cP A-pcP AD-pNT AD-cNT
E D-pN D-cN D-pcN
Note: S = Subject relations; E = Subject/environment relations;
A = Associated; D = Disassociated; AD = Asso- ciated/disassociated;
p = pragmatic mode; c = cognitive mode; pc = pragmatic/cognitive
mode; N = Nega- tive tone; P = Positive tone; NT = Neutral
tone.
ject. The code assigned to each program is listed in the "S"
(subject) row of each company in Table 3. Subjects were coded as
either D = disassociated, AD = associated/disas- sociated, or A =
associated. These terms are labels that identify the first of the
four narrative components in each program, the degree of unity
portrayed among all possible subjects of the letters. The following
illustrates how the con- tent from the text is used to render the
degree of subject as- sociation for the letters of companies E and
Z-77. Company E: Program I is about the state of the entire chem-
ical industry. Areas of strength and growth are identified for the
industry as a whole, which will be "cutting, fitting and sizing to
the realities of the marketplace." Toward the end of the program,
the CEO addresses Company E's position more specifically. Though
there are some indications of a combined set of company subjects,
(". . . our . .. businesses made im- portant contributions . . ."),
they are immediately followed by disassociative references to the
same parties (.1. . these businesses picked up healthy momentum . .
."). This was
287/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
coded as "D" in the subject row of Company E, Program 1. Program
11, in contrast, is short and company-oriented. Though the CEO at
times appears somewhat removed from the company ("increases in
costs-everywhere for every- thing"), there is intermittent
identification between CEO and company as subjects ("Company E has
been forecasting.... In our early planning"). Program 11 was thus
coded "AD." Program Ill surrounds Program I. The section preceding
Pro- gram 11 is short and Company E-oriented; the section fol-
lowing Program 11 is another lengthy and removed "voice of the
industry." The CEO associates more with the external environment
than with the company in both instances ("com- mitment to major
growth program" and "we are seeing the elimination of expansion"),
the "we" here seeming to include the entire industry. Subjects in
Program Ill were thus coded as "D." Company E's 1977 letter does
not discuss manage- ment changes, so Program IV is absent. Program
V again de- picts a distanced CEO who refers to ". . . the people
of Company E . . ." and was coded "D" for subjects. Company Z-77:
The subject is portrayed as an aggregated unit consisting of the
CEO and the entire company in Program I ("our results"), in both
parts of Program 11 ("our profit margins . . . our operations in
North America" and "we have been unable to offset"), and in both
parts of Program Ill ("we made progress" and "our problem"). The
subjects in each of these three programs were therefore coded "A."
In Programs IV and V, there is some ambivalence, with the CEO
identi- fying with the company in some instances ("we moved to-
wards our goal" and "our market and technological leadership"), yet
removed from the company at other times ("the company's operations"
and "the company's future prospects"). The subjects in these
programs were thus coded "AD." Subject/environment relations. The
question in subject/ environment relations is again the degree of
association, but this time the aim is to establish the spatial
domains of the narrative rather than to determine what combined
unit serves as subject. Companies' letters to shareholders usually
distin- guish between an internal and external environment. With
re- spect to their internal environments, CEOs usually wish letters
to be unambiguous in communicating growth, profit- ability, and
strength. External environments (including the chemical industry,
the national economy, and the world) are often more ambiguously
defined in relation to the company, with varying degrees of
company-environment association. The subject/environment relations
found in the letters were also coded on the basis of the use of
nouns and pronouns, this time in depicting internal and external
domains of the story. The relations noted in the letters of all ten
companies are listed in the "E" (environment) rows of Table 3. They
were also coded as either A = associated, AD = associated/
disassociated, or D = dissociated. The following summarizes how the
degree of subject/environment association was de- termined for the
letters of companies E and Z-77. Company E: A broad level of
association between the com- pany and the external environment is
portrayed throughout the letter. References to company and industry
are nearly in-
2881ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
distinguishable in Programs I and 111, "we" being the voice of
the entire industry ("As we start the new year, industry in-
ventories . . . are near normal levels."). Subject/environment
relations for Program I and Ill were thus coded "A." The ex- ternal
environment is not presented as a major force in any of the other
programs.
Company Z-77: The letter describes relations of the company with
the external environment only in terms of undesirable dependence on
external events that are perceived to be pri- marily negative.
Programs 1, 11, and Ill discuss results linked directly to
"weather-related problems," "economic condi- tions," "rise in ...
costs," and "weaker demand." Subject/ environment relations for
these programs were therefore coded "D." The external environment
is not present in Pro- grams IV or V.
The first two components identified in each of the narrative
programs indicate the existence of distinct internal and ex- ternal
relational differences in these two sets of letters. Se- miotics
seeks not only to uncover relational patterns but to make explicit
the rules that bind them and determine their function in a
narrative. Identification of the final two narrative components,
mode and tone, makes this possible. Mode. To qualify and categorize
the above relations so as to understand their function in the
story, semioticians must de- termine what Greimas (1966) called the
primary mode of each narrative program. Mode is a narrative
component that qua- lifies the relations between narrative forces
by distinguishing between those relations depicted within the
narrative's action sequences and those outside the action of the
plot. An action sequence is any part of the plot in which the
subject executes or receives an action. Relations taking place
through the plot's action sequences are said to develop the
subject's being- able-to-do and are designated as pragmatic;
relations that do not evolve through the plot's action are said to
develop the subject's knowing-how-to-do and are designated as
cognitive. For example, the relations developed in the letter may
be stated as part of the action of the plot, as in "division X de-
veloped a new product." The mode is pragmatic in that it in-
dicates a relation between narrative forces within an action
sequence. The sequence develops the competence or being- able-to-do
of the narrative subject in overcoming conflicts within the
narrative plot. However, if the primary function of a program is to
develop the subject's knowing-how-to-do, the mode of that program
is cognitive. These relations are not part of the plot's action
sequences. For example, a statement such as "the economy is weak"
signifies a state of mind of the narrative subject (not of the
author). It does not involve the subject's action but is
nonetheless a necessary step to- ward a resolution of the
narrative. Narrative mode is a critical indicator of narrative
structure. The patterned use of mode reflects underlying narrative
rules that govern how the different sets of relations noted above
may be combined to tell these companies' stories. If the narrative
subject is presented as a disaggregated group of subjects, the
narrative conflict can arise and be resolved through their
interactions. Since these take place within action sequences of the
story, the primary mode will be pragmatic. On the 289/ASQ, June
1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
other hand, the action sequences in a text portraying a unified
set of subjects are not likely to produce narrative conflict. Any
conflict in need of narrative resolution will develop outside of
the actions of this unified set of forces. The primary mode in this
case will be cognitive. Mode characterizations in this study's
narratives were coded on the basis of the relative emphasis of each
program on ac- tion (pragmatic) or state of mind (cognitive).
Programs with predominantly cognitive or pragmatic statements were
la- beled as such (p = pragmatic and c = cognitive). When there was
an intermingling of modes, the program was labeled pc =
pragmatic/cognitive, as shown in Table 3. The following summarizes
how the mode of relations in the letters of com- panies E and Z-77
was determined. Company E: Programs 1, 11, and III present an
intermingling of pragmatic and cognitive modes. A phrase like ". .
. lookout is cloudy" would be coded as cognitive, but the ". . .
industry will be cutting, fitting . . ." would be coded as
pragmatic. The first three programs in Table 3 were thus coded
"pc." Pro- gram V is a cognitive ("c") reference to the importance
of the company's people (". . . results will depend upon the
talents . . Company Z-77: The shortest program, Program 1, is
pragmatic ("p") when discussing subject relationships
("Agricultural chemicals made an important contribution . . .").
Program 11 presents cognitive relations in describing reasons for
the company's current position ("c"). Program III continues to be
mostly cognitive, with the striking exception of the first part of
Program 111, a pragmatic textual block outlining the means of
actively distancing the company from hindering forces ("pc").
Relations in Program IV are pragmatic ("p") ("we moved ... towards
our goal"). The letter ends on a cognitive note ("c") ("proven
capabilities . . . should give us all confi- dence"). Tone. The
final narrative component, tone, qualifies the rela- tions noted
above among the various textual forces as either negative or
positive. The narrative tone signifies the attitudinal orientations
of the subjects within the various relationships. A negative tone
denotes that the relation is a hindrance to the subject's quest; a
positive tone denotes that the relation is a help. Tone is a
critical indicator of narrative structure. The patterned use of
tone, like mode, signifies the rules by which the various types of
relations among textual forces may be com- bined to create the
image that is the object in these letters. If the narrative's
primary source of conflict is developed within highly associated
internal and external environments, the tone by which both the
internal and external relations are depicted will be predominantly
positive. Depicting them in negative in- terrelationships would
counter the ultimate object of the letters. In the other case, the
primary source of narrative con- flict is developed between an
internal subject set and an ex- ternal set of hindrances. Here, the
relations developed to resolve the conflict in the narrative will
be between a posi- tively depicted set of subjects and a negatively
depicted set of hindrances. Tone characterizations in this study's
narratives were coded by distinguishing between positive and
negative descriptions 290/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
of both actions (pragmatic) and causal attributions of events
(cognitive). Clear indications of a negative (e.g., "bad weather")
or a positive (e.g., "tremendous growth") tone were labeled as
such, P = positive, N = negative. The lack of a clear positive or
negative tone was labeled as NT = neutral. The tone
characterizations for the companies in the study are shown in Table
3. The following summarizes how tone was coded for the letters of
companies E and Z-77. Company E: When referring to the narrative
subjects, the tone of Programs I and III is mixed ("NT"); in
Program 11 it is primarily positive ("P") ("enjoy increases in
sales . . . good part of benefit from gains . .. cannot be
converted into gains in earnings" rather than "losses due to").
When referring to the environment, the tone of Program I is
positive ("P") ("U.S. demand ... a pleasant surprise"); the tone of
Program III is negative ("N") ("costly regulatory program").
Program V is a neutral ("NT") reference to both the difficulties
and the achievements of the previous year. Company Z-77: Program I
is positive ("P") when discussing subject relationships ("important
contribution"), but environ- mental relationships in the same
program are depicted as negative ("N") ("weather-related
problems"). In Program 11 all internal problems are expressed as
outcomes of external conditions. Subject relationships were coded
as positive ("P") and environment as negative ("N"). Program III is
a continua- tion of the same, as the company is portrayed as
actively re- moving itself from negative external influences.
Subject relationships in Programs IV and V are depicted as neutral
("NT"). The second step of Phase 2 thus served to identify the four
components in each program and to classify each component within an
oppositional framework. The patterns of opposi- tional combinations
of these components point to narrative rules that determine how the
components may be meaning- fully combined in this narrative system.
For example, when the subjects are associated in these narratives,
the mode is most often cognitive. The following is a summary of the
pat- terns identified above. The two-way arrows indicate the pre-
dominant interactions: Associated subjects
-
Table 4
Interactive Narrative Components*
Programs Components Coding I II III IV V
Joint-venturing firms
Subject Associated 0 0 0 0 0 Disassociated 5 3 3 2 4
Subject/environment Associated 2 3 3 0 1 Disassociated 0 0 0 0
0
Mode Pragmatic 4 2 4 3 4 Cognitive 0 1 0 0 0
Tone External Positive 4 2 2 0 2
Negative 0 0 1 0 0 Internal Positive 0 1 0 2 0
Negative 0 0 0 0 0
Non-joint-venturing firms Subject Associated 3 2 2 2 2
Disassociated 0 2 0 0 0 Subject/environment Associated 0 0 0 0
0
Disassociated 2 5 4 0 2 Mode Pragmatic 4 0 2 4 2
Cognitive 1 4 0 1 1 Tone
External Positive 0 0 0 0 0 Negative 1 3 1 0 1
Internal Positive 5 1 3 2 0 Negative 0 1 0 0 0
* Numbers are the number of companies whose letters reflect the
component in each of the two opposing forms. The 1979 letters of
companies Y and Z are not included in this table.
closely linked with their environment and relating to that en-
vironment in a positive manner within action sequences. Firms with
no joint ventures also begin their stories with a discussion of
company strength and growth. Theirs, however, is predominantly a
narrative pattern depicting an aggregated group of subjects, shown
as distinct from their environment, the positive tone relating to
the internal subject group. A curious inconsistency with the
structural pattern identified earlier for the whole textual system
is that the mode of Pro- gram I in this second group of stories is
primarily pragmatic. According to the patterns indicated earlier,
conflict will not be established within the action of the plot
(pragmatic mode) when the subjects involved in that action are
presented as a single force. How then is the narrative conflict
established in the next program, the program relating to subject
failures? The answer lies in the textual transformation that occurs
in Program 11 (second column in Table 4). Whereas joint-ven- turing
firms continue the narrative pattern of the previous program, firms
with no joint ventures tell the story differently. Subject
association and internal positive tone, both used to set the stage
for the story (Program 1), are not used in Pro- gram 11. Moreover,
the mode switches from pragmatic to cognitive, moving the focal
relations outside of the subjects' actions. To explain the
transformation occurring in Program 11 in the second group of
letters above requires examination of the re-
292/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
suits listed in Table 4 in a two-way interaction, across each
row (dynamic relations across a text, referred to as syntag- matic
relations) as well as down each column (static relations among
possible components of a single program, referred to as
paradigmatic relations). Figure 3 illustrates the sequential path
of each narrative component across the five programs. The
horizontal axis shows the sequence of programs; the vertical axis
shows the number of firms whose letters reflect the form of
narrative component depicted in the graph. The graph illustrates
the extent to which the configuration of components in each program
conforms to the narrative rules derived earlier. It highlights the
consistency of pattern in the first group of letters and the
deviations of the second group.
Figure 3. Narrative-component sequential paths.*
Joint-Venturing Firms
5 ------- Pragmatic Mode * s . _ Subject
E 4 Disassociation LL " \, " "External
A 3 Positive Tone -0 ~~~~~~~~'\ .' / ~~~~~~~~~Subject!
E 2 ' \ ~t / _ bEnvironmental 1Z" ,.' ~ _ \ / Association
s \ // I I I s~
I 11 111 IV V Programs
Non-Joint-Venturing Firms Subject
5 - Association / o ~Subject/
E 4 - / - Environmental LL / , , Disassociaion
0 3 -- , , \ ~----- Cognitive Mode -n Ad / s \External E 2 -
Negative Tone
I II 111 IV V Programs
'The graphs are based on the data in Table 4. The numbers are
taken directly from the table when the set of letters consistently
used one form of the narrative component. When both forms of the
component were present, the number used on the graph is the
difference be- tween the number of firms using one form and its
opposite.
Figure 3 illustrates two important aspects of the use of the
four narrative components across the five programs of a letter.
First, the sequential paths illustrate the paradigmatic links
between the numbers in each column of Table 4. The extent to which
the four paths follow each other indicates in- terdependence of the
components. For example, the paral- lelism in the componential
structure of the first group of letters marks a strong pattern:
subject disassociation, sub- ject/environment association,
pragmatic mode and externally directed positive tone are used as
narrative devices in close and consistent interaction. When a large
number of letters in 293/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
the first group uses one of these components, a similarly large
number of letters uses the others. In the second group of letters,
subject/environment disassociation, external nega- tive tone, and
cognitive mode tend to follow the same pat- tern, whereas the
subject association line moves in opposition. The four components
do not appear to be equally interdependent, the deviating paths
being particularly salient in Program 11. Second, the sequential
paths of Figure 3 illustrate the syntag- matic links between the
numbers in each row of Table 4. The relatively high and low points
of the graph indicate the sa- lience of the given configuration of
narrative components in different parts of the letters. The highest
points for all four components in the first set of letters are at
the beginning and end of the narratives, indicating that this
particular set of nar- rative devices was most important at these
two points for the successful development of their stories. The
graph of the second group of letters shows that both the highest
and lowest point is in Program 11, indicating that the set of
narra- tive devices was both most important and most inconsistent
at this point. The implications of this narrative configuration are
discussed later. Narrative transformations may also occur across
letters from the same company from year to year. To allow for such
transformations, the present analysis was carried out over a
three-year period, 1977-1979. The patterns revealed are stable (and
thus were consolidated in the results shown in Table 3) across the
three-year period for all but two of the companies, Company Y and
the non-joint-venturing company used as an illustration in this
paper, Company Z. An inter- esting shift occurs in the text of
these two non-joint-venturing firms. The 1977 text in both cases
follows the pattern of the other non-joint-venturing firms. After
1977, the pattern changes to a near turnaround, following the
narrative struc- tures of the joint-venturing firms of this sample.
Figure 4 con- trasts the 1977 and 1979 letters. It illustrates the
consistency of narrative pattern within each year's letter and the
shift that occurred over time. In both cases, this shift occurred
at the time that a new CEO took over. Though it is impossible to
state definitively that the change in leader caused a shift in
company posture (along the dimensions examined in this study), it
appears to be more than a coincidence that the eight companies with
no change in leadership followed the same patterns across the three
years and that the two companies whose leaders changed did not
follow the same patterns. This observation provides sup- port for
the premise that the CEO plays an important role in determining the
kind of story that is told in portraying a posi- tive company
image. Deep Structure In Phase 2 of the analysis the narrative
structures were iden- tified within each program in the two sets of
letters. The im- plications of both the structural regularities and
irregularities in the letters are best understood when expressed
within the context of a logical oppositional structure underlying
the en- tire set of letters. The schema used to depict these
opposi- tions is the deep-structure schema identified earlier.
294/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
Figure 4. Shift of pattern of companies Y and Z. Cognitive
Mode
Subject Association
Joint-Venturing Firms Subject/Environmental 1977
Disassociation
External Negative Tone
.I- 2 0
4 1 11111 IV V
Programs
Subject Disassociation-
Pragmatic Mode Non-Joint Venturing Firms
1979 Association
External Positive Tone
tS t S/t < \
.I-
E1
/ ~ ~ ~~~X
I II III IV V Programs
The graphs are based on the data in Table 4. The numbers are
taken directly from the table when the set of letters consistently
used one form of the narrative component. When both forms of the
component were present, the number used on the graph is the
difference be- tween the number of firms using one form and its
opposite.
The complete deep-structure schema or "semiotic square" (Greimas
and Rastier, 1968) of a set of oppositions underlying these letters
is presented in Figure 5. Also presented is a summary list of the
narrative rules or conventions that dictate how narrative units may
combine to generate underlying pat- terns.
The stories of joint-venturing firms consistently follow the
narrative rules that allow a simultaneous presentation of weak
external boundaries and the logically implied strong internal
boundaries (Position 1). There is no deviation to values on the
right side of the square, indicating a rather straightforward
application of the narrative rules. What happens in the other set
of narratives is somewhat more interesting. The stories begin by
depicting subject/ environment disassociation and subject
association, narrative components signaling values of strong
external and weak in- ternal boundaries. Yet the pragmatic rather
than the cognitive mode predominates. The pragmatic mode generally
charac- terizes relations developing in a disassociated set of
subjects' action sequences, whereas the subjects here are presented
as a unified force. The use of the pragmatic mode suggests the
beginning of a tension between strong and weak internal 295/ASQ,
June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Figure 5. Deep structure of shareholder letters with
corresponding narrative rules'
Weak JOINT-VENTURING FIRMS Weak External 1 Internal Boundaries
Boundaries (E: A-P) (S: A-c)
3 5 2
Strong \ Strong Internal External Boundaries 4 Boundaries (S:
D-p) (E: D-N)
Strong NON-JOINT-VENTURING FIRMS Strong External 3 Internal
Boundaries Boundaries (E: D-N) 7 (S: D-p)
1 5 2
Weak Weak Internal External Boundaries 4 Boundaries (S: A-c) (E:
A-P)
Numbers on the axes refer to internal and external boundary
combi- nations. S = Subject relations; E = Subject/environment
relations; A = Associated; D = Disassociated; p = pragmatic mode; c
= cogni- tive mode; N = Negative tone; P = Positive tone.
boundary definitions. Program I is thus located on the contra-
dictory axis of Position 5. This tension caused by the
contradiction is resolved in Pro- gram 11 as the narrative
structures more fully take on the sub- ject disassociation
characteristic of the underlying value of strong internal
boundaries. A tension between complex con- traries is developed,
since the narrative structures now simul- taneously reflect the
dominant positive value and its contrary, strong external and
internal boundaries (Position 3). This posi- tion allows the
narrative conflict to be established unambigu- ously as outside of
the company's sphere of action, while acknowledging the internal
discord that would accompany such conflict by de-emphasizing
subject association and in- ternal positive tone. The cognitive
mode and intensified sub- ject/environment disassociation in
Program 11 prevent the narratives from moving to Position 2 of the
square. In subse- quent programs, the narratives return toward
Position 5, the subject/environment disassociation remaining
consistent (ex- cept in Program IV, where the external environment
is largely absent from the story) and the other three components
causing the story to vacillate between values of weak and strong
internal boundaries. 296/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
The final underlying deep structure identified in the stories of
the joint-venturing firms is consistently that of Position 1 in
Figure 5. Moreover, this deep structure is developed through the
consistent and patterned application of narrative rules. The
stories of the group of firms without joint ventures, how- ever,
display underlying consistency only in embracing the dominant value
of strong external boundaries. Internal bound- aries are variably
defined as weak or strong, thereby pointing to a contradictory deep
structure identified as Position 5 in the semiotic square (with
temporary movement to Position 3 in order to develop the narrative
conflict). Though the under- lying deep structure of this group of
texts is in opposition to that of the joint-venturing group, the
opposition is consistently contradictory only in terms of external
boundary definitions.
DISCUSSION The analysis moved through successive phases of
breaking a set of thirty narratives into component parts,
characterizing each part, and then identifying patterns in the way
the parts are combined. The final analytic phase offered a map of
the texts' underlying structures, their positions in the semiotic
square. The results shown in the semiotic squares illustrate both
the interplay of oppositional structures within a single set of
narratives and the differences in the oppositional struc- tures
between the two sets. The interplay of logical structures within
the set of letters of joint-venturing firms highlights the
patterned simplicity under- lying- a group of highly variable
stories, much like the recurring patterns previously found in
multiple myths and folktales (Propp, 1958; L6vi-Strauss, 1976). The
interplay of structures within the letters of non-joint-venturing
firms is more com- plex, with logical inconsistencies underlying
the narrative. The results of the analysis of oppositions between
the two sets of narratives partially support this study's
proposition. The letters of the five joint-venturing firms
consistently por- tray weak external and strong internal
boundaries. The letters of the non-joint-venturing firms
consistently portray strong external boundaries but do not portray
a consistent definition of internal boundaries. There are at least
two alternative means of interpreting the inconsistent logic found
in the letters of the latter group. One is to maintain that the
contexts of these two sets of letters are similar and that the
rules governing narrative consistency apply equally to both. From
this perspective, the letters of the non-joint-venturing group of
firms present an inherently un- stable image of the firms and their
internal relations. The contradictory narrative structures relating
to the strength of internal boundaries within this group of letters
suggest the lack of a coherent set of shared meanings about the
extent of unity or separateness of internal company units. The
environmental context within which these letters were written lends
support for this first interpretation. Each of the firms in the
sample was heavily involved in the stagnating in- dustrial
chemicals business and felt pressure to retool and move into the
more rapidly growing specialty chemicals markets. The specialized
requirements of these new busi- nesses suggest the need for
increased autonomy and the 297/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
separateness of the internal units of a company. The under-
lying contradictions regarding internal boundaries of one set of
letters may thus reflect the transitional instability that might
accompany such a change. Future research efforts could usefully
extend the comparison between the two groups of firms to determine
whether, in fact, the results of this study reflect differing rates
of institutionalizing change. Alternatively, the inconsistencies of
the letters of non-joint- venturing firms may be interpreted as
suggesting that the rules that apply to narratives such as myths
and folktales do not govern narratives in all contexts.
According-to this perspec- tive, the objective constraints
surrounding the development of these corporate messages invalidate
the narrative rules of logical consistency. Unlike myths and
folktales, letters to shareholders are bound by a set of external
imperatives. Even if theirs is a story of a highly unified set of
narrative subjects, they are required to communicate a certain
number of activi- ties of separate units of the firm. Given this
interpretation, the challenge for future research is to examine a
broader array of narratives, bound by diverse external constraints,
in order to ascertain the extent to which these govern the way the
story is told.
CONCLUSIONS The primary contribution of this study lies in its
presentation of semiotics as a method for analyzing how signs
combine in a patterned way to convey meanings within a particular
con- text. Unlike more traditional forms of content analysis (e.g.,
Bettman and Weitz, 1983; Staw, McKechnie, and Puffer, 1983; Bowman,
1984; Salancik and Meindl, 1984), semiotics relies on an a priori
value schema and interprets each piece of data only in light of
this schema. Traditional content analysis focuses on individual
statements and actions, assuming that "annual report discussion,
line-by-line, is a reasonable surro- gate for real activity"
(Bowman, 1984:64). In his study of risky behaviors of firms, for
example, Bowman (1984) focused on .the variants of specific
activities (extent of acquisition, litiga- tion, and new-venture
involvement of the firm), without at- tempting to explain the
commonalities that bind these behaviors. In contrast, semiotics
emphasizes the underlying values that differentiate the meaning of
behaviors. The underlying oppo- sitions defining a belief system
related to risky behaviors, for instance, might be "loss," "gain,"
"non-loss," "non-gain." In such a study, the aim of a semiotic
analysis would be to pro- vide the framework that defines the
bounds of possible be- haviors within a particular context and to
identify the rules that account for patterns in that behavior. It
would not at- tempt to explain the multiple variations of the
behaviors themselves or their performance implications. Semiotics
exposes underlying oppositional values through the identification
of structural invariants. As illustrated in this study, the concern
is with structures embedded in a system of signs (in this case, a
series of stories), rather than struc- tures of the intentions of
the sign-sender (author of the letters). The abstract concern with
formal oppositions has been criticized for its attention to purely
cognitive representa- tions lacking rich descriptions of behaviors
and feelings
298/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
(Denzin, 1983; Manning, 1987) and for its inattention to the
understanding brought to the message by the reader and the sender
(Peirce, 1940). These criticisms are useful in defining and
delimiting the potential role of semiotics in organizational
research. To discover an invariant does not always mean to discover
a truth. It does, however, identify problem areas for future
research in systematic and differentiable terms. The links that
were uncovered in this study between boundary definitions in
organizational messages and new- venture structures are exploratory
and certainly suggestive. The value of such a study lies not in
what it "proves" but in what it exposes for further exploration and
questioning. Though researchers have long been intuitively aware of
the probable links between organizational beliefs and strategic
behavior, it has been difficult to capture the apparent com-
plexity of these links with traditional research methods (Broms and
Gahmberg, 1983). The semiotics of CEOs' letters calls attention to
the simple belief patterns that may underlie the seemingly complex
relationships between organizations and their environments.
REFERENCES Barley, Stephen R. 1983 "Semiotics and the study
of
occupational and organizational cultures." Administrative
Science Quarterly, 28: 393-413.
Berg, Sanford, Jerome Duncan, and Philip Friedman 1982 Joint
Venture Strategies and
Corporate Innovation. Cam- bridge, MA: Oelgeschlager, Gunn &
Hain.
Bettman, James R., and Barton A. Weitz 1983 "Attributions in the
board
room: Causal reasoning in cor- porate annual reports." Ad-
ministrative Science Quarterly, 28: 165-183.
Bowman, Edward H. 1984 "Content analysis of annual
reports for corporate strategy and risk." Interfaces, 14:
61-71.
Broms, Henri, and Henrik Gahmberg 1983 "Communication to self in
or-
ganizations and cultures." Ad- ministrative Science Quarterly,
28: 482-495.
Denzin, Norman K. 1983 "Towards an interpretation of
semiotics and history." Paper presented at the Colloquium on
Criticism and Interpretive Theory, University of Illinois,
Urbana-Champaign.
Eco, Umberto 1979 A Theory of Semiotics. Bloom-
ington, IN: University of In- diana Press.
Greimas, Algirdas J. 1966 Semantique structurale. Paris:
Larousse. 1987 On Meaning: Selected
Writings in Semiotic Theory. Minneapolis: University of
Minnesota Press.
Greimas, Algirdas J., and Frangois Rastier 1968 "The interaction
of semiotic
constraints." Yale French Studies: Game, Play, Litera- ture. New
Haven, CT: Eastern Press.
Haidu, Peter 1982 "Semiotics and history." Se-
miotica, 40 (3/4): 187-228. Harrigan, Kathryn R. 1985 Strategies
for Joint Ventures.
Lexington, MA: Lexington Books.
Levi-Strauss, Claude 1963 Structural Anthropology, vol. 1.
New York: Basic Books. 1976 Structural Anthropology, vol. 2.
New York: Basic Books. Manning, Peter K. 1987 Semiotics and
Fieldwork.
Newbury Park, CA: Sage.
Miller, James C. 1979 "The psychology of innovation
in an industrial setting." In W. G. Lawrence (ed.), Ex- ploring
Individual and Organi- zational Boundaries: 205-216. New York:
Wiley.
Miller, Eric J., and Albert K. Rice 1967 Systems of
Organization. New
York: Tavistock. Peirce, Charles Sanders 1940 Selected Writings
of Charles
Sanders Peirce. London: Har- court, Brace.
Propp, Vladimir 1958 Morphology of the Folktale.
Austin, TX: University of Texas Press.
Salancik, Gerald R., and James R. Meindl 1984 "Corporate
attributions as illu-
sions of management control." Administrative Science Quar-
terly, 29: 238-254.
Saussure, Ferdinand de 1916 Cours de linguistique g6n6rale.
Paris: Payot. Staw, Barry M., Pamela I. McKechnie, and Sheila M.
Puffer 1983 "The justification of organiza-
tional performance." Adminis- trative Science Quarterly, 28:
582-600.
Wilden, Anthony 1980 System and Structure: Essays
in Communication and Ex- change. New York: Tavistock.
299/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
APPENDIX: Samples of Letters to Shareholders, with Program
Codings for Surface Structure
Company E (Joint-Venturing Company) 1977 Letter to Shareholders
To Our Shareholders:
[Program l: Implied goal: Show areas of strength and growth]
[Company E] earned $108.2 million, or $6.09 per primary share in
fiscal 1977. These earnings represent a 16.7% return on
shareholders' equity and a 12.1% return on invested capital-a
respectable showing in a year in which the economic environments in
our major markets ranged from surprisingly strong to unexpectedly
depressed. Looking back: 1977 trailed 1975 and 1976, two years of
extraordinary profit- ability for the [R] industry. The $6.09 per
share was off 38% from 1975 and 21% from 1976. Down is down, but
1977's 16.7% return on shareholders' equity is a good statement of
the company's basic earnings power, and it is a fine base upon
which to build. The [R] industry is still wrestling with excess
capacity, but the growing into new plant is less painful than might
have been expected. U.S. demand-a pleasant surprise-totalled 51
million tons, up 3% from 1976, which in turn was up 16% from 1975.
The industry's [P] shipments increased an impressive 22%. While
some [P] remained in the U.S. distribution system at yearend,
product going on the ground increased by approximately 6%. In [H],
total shipments increased 10%, with a notable 12% increase in
exports. As we start the new year, industry inventories of [H] and
[P] are near normal levels and sales possibilities are favorable.
With energy, labor, and other vari- able costs on the rise,
[Company E] plans to continue its policy of tight in- ventory
control and stringent controls on extension of credit. Quite
clearly, prudent money management is the key to good results in the
[R] business. The outlook for [A] is cloudy. Excess capacity
exists, and the industry still has some new plant under
construction. Accordingly, over the next 12 to 18 months the
industry will be cutting, fitting, and sizing to the realities of
the marketplace. The availability of natural gas is becoming
increasingly unpre- dictable. Shortages could markedly reduce
output. For [Company E's] part, we control sufficient natural gas
for our require- ments, our [A] costs are highly competitive, and
we are reasonably certain of sales volumes close to capacity.
Prices are apt to be unsatisfactory, at least through fiscal 1978.
As predicted, our chemical and industrial businesses made important
contri- butions to fiscal 1977 operating earnings-39% of the total.
Selling into markets highly influenced by general economic
conditions, these businesses picked up healthy momentum in the
United States as the year progressed. In Europe, important markets
for [Company E] industrial products were de- pressed, and a change
for the better is not in sight. Japanese business moved sidewise
last year and is likely to do the same this year. [Program Ill:
Implied goal: Justify expenditures] Capital spending totaled $175
million in fiscal 1977, the fourth year of com- mitment to a major
growth program. As of June 30, 1977, something over 50% of the
approximately $750 million invested over the past four years was
not yet earning the company's objective of a 12% return on invested
capital. Most of these new assets will earn up to plan in due
course. The pruning of marginal assets continues, and those without
potential will be sold or liquidated. [Program II: Implied goal:
Show reasons for poor performance] [Company E] has been forecasting
earnings at its Annual Meeting each Oc- tober, and the practice
will be followed again this year. In our early planning, it is
clear that we have a good chance to enjoy increases in sales almost
across the board. Less predictably, but almost inexorably, we will
experience very large increases in costs-everywhere for everything.
Increases in prices seem likely, in the short run, to cover
approximately one-quarter of the anticipated increases in costs. It
follows, therefore, that a good part of the benefit from gains in
volume cannot be converted into gains in earnings per share in
fiscal 1978.
300/ASQ, J une 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
[Program III continued: Implied goal: Justify expenditures] On
the matter of increases in costs, it should be noted that [Company
E's] environmental protection programs will total approximately $36
million in fiscal 1978. While environmental protection is the most
costly of the regula- tory programs from our point of view, the
array of federal regulations-EPA, ERISA, OSHA, WWO, SEC, et al-is
becoming increasingly formidable. Their costs have to flow through
to price sooner or later. This is not the proper place to discuss
the complex issues involved in gov- ernmental regulation of
business activity, but it does seem necessary to voice an appeal
for thoughtful consideration of costs vis-6-vis the public ben-
efit derived. Clearly, in some areas the rules seem to be
increasing the costs of products to the producers and to the
consuming public unnecessarily. Also, we are seeing the elimination
of expansions that might have created new activities and new jobs.
The company has been seeking a [Product C] producer as a logical
extension of our commitment to natural resource businesses.
Accordingly, we are pleased to have acquired a [Product C] product
in [the south]. This property should give us a good start in the
business. We continue to think that, sooner or later, the [P]
industry and the Saskatch- ewan government will be able to meet on
common ground. The government has moved to correct some past
injustices, and these conciliatory gestures, plus some recent
meetings with officials, give us hope that 1978 will be a year of
progress in this difficult relationship. Following past practice,
this Annual Report features one of a variety of re- sponses to
corporate social responsibilities. During fiscal 1977, a major
effort to improve the company's safety performance was launched. We
are pleased to note that lost-time accidents were reduced by 43%.
The program is de- scribed on page 10. [Program V: Implied goal:
Show future outlook] Fiscal 1977 was not an easy year; 1978 starts
nicely, but again, it will not be easy. As always, the results will
depend upon the talents of all people-in all jobs. On behalf of the
Directors and the Executive Office, it is my privilege to thank the
people of [Company E] for what they have achieved and for their
continuing commitment to the company goals. Company Z
(Non-Joint-Venturing Company) 1977 Letter to Shareholders To the
Stockholders of Company Z: [Program I: Implied goal: Show areas of
strength and growth] Sales from continuing operations in 1977 were
$1,123,865,000, an increase of 11 % over 1976. Net earnings were
$45,624,000-down 1 % from re- stated results for the prior year.
Agricultural chemicals made an important contribution to our
results as prof- itability improved following the weather-related
problems of 1976. Profits for industrial chemicals also were up,
reflecting the generally better economic conditions. [Program II:
Implied goal: Show reasons for poor performance] Profitability of
our [XI business declined because of (1) a sharp drop in U.S. paint
sales due to extremely cold weather in early 1977, (2) increased
manu- facturing costs for acrylate monomers in North America caused
by the need to operate the older acetylene process during the
proving-in period for the large new propylene-based plant, and (3)
sharply reduced [XI sales in Latin America due to depressed
economic conditions there. The [YM business segment was hurt by
overcapacity in the acrylic sheet in- dustry in the U.S., and this
situation was aggravated by imports of acrylic sheet. Downward
price pressure eroded our profit margins, and this will continue
into 1978. These problems in [XI and [YI affected our operations in
North America and Latin America, and profits in these two regions
were down. Increased earnings were reported by both the Pacific and
European regions. [Program Ill: Implied goal: Justify expenditures]
Divestiture of [PI and [Q] Products- We made substantial progress
in 1977 towards achieving our objectives of reducing debt,
strengthening the balance sheet and investing only in those
businesses in which we are-or can expect to become-a primary
compet- itor. To these ends: We sold our [P1 operations in the U.S.
and continued our efforts to dispose of our [P1 business in Brazil.
We sold our [01 business in 301/ASQ, June 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
the U.S. and actively pursued leads to sell our worldwide [Q]
business. We decreased our debt to equity ratio from 81 % to a far
more satisfactory 61 % at yearend 1977. On December 29 we sold our
[PI plant for $55 million, thereby completing the divestiture of
our domestic [PI business. We announced in October an increase in
our provision for loss on disposal of [PI from $40 million to $57
mil- lion. Part of this larger provision was due to the relatively
low sale price of the [P1 plant, but the bulk of the additional
writedown was related to Brazil where the [PI industry remains
depressed. The value now assigned to these assets in Brazil is
approximately $30 million. During 1977 we decided to divest our [Q]
businesses and to account for these operations as a discontinued
business segment. This decision came after an extensive review of
our position and of the long and expensive pro- cess now required
to bring new products to the marketplace. We decided our resources
could be used more efficiently in our [RI businesses. We sold
[sudsidiary SI in September, and the aftertax profit on this
sale-$16,440,000-was recorded as a gain from the disposal of a
discon- tinued operation. At yearend our other [Q] business had not
yet been sold. Sale of our Brazilian [P1 assets and our worldwide
[Q] business, plus internally generated cash flow, will permit
further debt reduction during 1978. We ex- pect to reach our
corporate target of 50% debt to equity ratio during the course of
the year. Now that we have achieved a strong balance sheet, we are
committed to maintaining that strength and to further improvement.
[Program IV: Implied goal: Show management changes] Board of
Directors- We moved forward in 1977 towards our goal of
strengthening the board and closely involving the outside directors
in corporate strategy and planning. In April, William L. Mobraaten,
president of The Bell Telephone Company of Pennsylvania, was
elected to the board. John H. McArthur, associate dean of the
Harvard Business School, was elected to the board in July. The
newly-formed Finance and Strategic Planning Committee of the board
met regularly during the year to review the company's operations
and its fi- nancial and strategic plans. [Program II continued:
Implied goal: Show reasons for poor perfor- mance] Outlook for the
Future- Inflation has been a major challenge for a number of years.
Since 1973 we have been unable to offset fully the rapid rise in
raw materials, energy and other costs. This inability to maintain
an adequate relationship between price increases and cost increases
has been the principal reason for the decline in our gross profit
margin from 33.8% in 1973 to 27.4% in 1977. [Program IlIl
continued: Implied goal: Justify expenditures] A major portion of
our problem has been the rapid escalation of the cost of natural
gas which we use in the U.S. for feedstock and fuel purposes. In
1973, we used about 30 billion cubic feet (bcf) of natural gas
annually at a cost of about $6 million. By the beginning of 1977 we
had reduced our an- nual usage to about 20 bcf, but this amount
cost almost $40 million. Further economies will reduce natural gas
usage to about 12 bcf in 1978. These changes will involve some
increases in other costs, but the net savings to the company will
be about $12 million annually-starting in 1978. Our new raw
material base gives us a more stable cost structure. If our pro-
duction in the last two years had been based on our 1978 production
pro- cesses, our raw material cost increases in North America would
have been 1.2% in each year instead of 5%. Although we anticipate
increases in 1978 will exceed 1.2%, we expect to achieve a much
better balance of price in- creases to cost increases in the
future. In the early 1970s we anticipated that economic growth in
the middle and latter part of the decade would be greater than is
now likely. We based our construction program on these earlier
growth estimates and, as a result, now have excess manufacturing
capacity. This excess capacity penalizes current earnings, but it
will allow us to expand in the future at lower than normal capital
cost. [Program N/: Implied goal: Show future outlook] Our company
is better positioned for profitable growth than we have been for
some time. The existing line of products, more stable raw material
cost structure and efficient installed plant capacity will
reinforce our market and technological leadership. Our long-term
future is dependent upon our re-
302/ASQ, J une 1989
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
-
Organizational Boundaries
search effort's maintaining the competitiveness of the current
product line and creating new proprietary products. In this report
a number of 1977 product introductions are presented. These
substantiate again our ability to recognize and match product needs
in many diverse markets with product developments in various fields
of chemistry. The proven capabilities of [Company Z's] experienced
team of scientists, en- gineers and other dedicated employees
should give us all confidence in the company's future
prospects.
This content downloaded from 181.193.19.161 on Tue, 18 Jun 2013
16:19:08 PMAll use subject to JSTOR Terms and Conditions
Article
Contentsp.277p.278p.279p.280p.281p.282p.283p.284p.285p.286p.287p.288p.289p.290p.291p.292p.293p.294p.295p.296p.297p.298p.299p.300p.301p.302p.303
Issue Table of ContentsAdministrative Science Quarterly, Vol.
34, No. 2 (Jun., 1989), pp. 169-347Front MatterThe Decoupling of
CEO Pay and Performance: An Agency Theory Perspective
[pp.169-189]The Lid on the Garbage Can: Institutional Constraints
on Decision Making in the Technical Core of College-Text Publishers
[pp.190-207]Whether and When? Probability and Timing of Incumbents'
Entry into Emerging Industrial Subfields [pp.208-230]Power from
What? A Reexamination of Its Relationships with Structural
Conditions [pp.231-251]Managing to Be Fair: An Exploration of
Values, Motives, and Leadership [pp.252-276]A Semiotic Analysis of
Corporate Language: Organizational Boundaries and Joint Venturing
[pp.277-303]About the Authors [pp.304-305]News and Notes
[p.306]Book ReviewsOther Reviews [pp.321-324]Other Reviews
[pp.325-326]Other Reviews [pp.327-330]Other Reviews
[pp.330-332]Other Reviews [pp.332-335]Other Reviews
[pp.335-338]Other Reviews [pp.338-341]Other Reviews
[pp.341-342]
4 Reviews on Ethnography, Work, and Organizationuntitled
[pp.307-311]untitled [pp.311-315]untitled [pp.315-318]untitled
[pp.318-321]
Publications Received [pp.343-345]Back Matter [pp.346-347]