ON CONCEPTUALIZATION AS A SYSTEMATIC PROCESS

A. J. J. van Breemen and Janos J. Sarbo

Institute for Computing and Information Sciences, Radboud University, Nijmegen, The Netherlands

Keywords:

Conceptualization, signs, Peirce, semiotic sheet.

Abstract:

We are concerned with the early phases of ER-modeling consisting in the primary conceptualization of the

underlying application domain. To this end we introduce a process model for the generation of a domain

description. In virtue of its close relation with cognitive activity, this model enables the modeler as well as the

user to comprehend the concepts of the resulting domain in a natural way.

1 INTRODUCTION

Is it possible to classify the concepts ‘generated’ dur-

ing modeling processes in types? By knowing the

kinds of concepts that can be distinguished, we could

make conceptualization more systematic which in

turn may improve its effectiveness (less errors) and

conceptual efﬁciency (Liu, 2000). To this end we in-

troduce a process model for conceptualization. Our

model can be derived on the basis of an analysis of the

properties of cognitive activity, as it has been shown

at last year ICEIS conference (Sarbo, 2007); and on

the basis of a semiotic analysis of representation, as

it will be developed in this paper. Whereas the ﬁrst

approach may provide more insight in the computa-

tional aspects of conceptualization as a process, the

second one opens the way towards using the types of

conceptualization events as pigeonholes during prob-

lem speciﬁcation.

In past research we have shown that our pro-

cess model enables uniform representation in differ-

ent knowledge domains. This is why our approach

may be interesting for organizations involved in mod-

eling processes. The aim of such processes is to make

explicit and reconcile knowledge obtained from dif-

ferent stake holders, in a (semi-)formal model. The

knowledge involved is generated with the aim of real-

izing goals in individual conceptualization processes

that themselves serve a common goal: the ‘genera-

tion’ of a shared, common understanding of a prob-

lem. In virtue of this common goal, conceptualiza-

tion in organizations offers an excellent occasion to

execute control.

The knowledge involved in each individual busi-

ness process is tightly knit with the goal it serves to

realize. This shapes the attitude of the stake holders

of different domains. It is only natural that, as a rule,

they will understand and value other processes from

their own habitual point of view. Reconciliation starts

with the decision to consciously subsume knowledge

needed for the different processes under the common

goal. But this only expresses the intention. In or-

der to realize the intention, it is necessary to explicate

the domain speciﬁc knowledge in such a way that the

stake holders of related domains are able to under-

stand and value what is going on. This is where the

uniform modeling potential of our approach may be-

come useful. For, in that case the different stake hold-

ers get acquainted with the form in which knowledge

is conceptualized when their knowledge processes are

modeled. This familiarity with the form comes in

handy when stake holders of different domains par-

take in Information System (IS) building. It is here

that our process model for systematic conceptualiza-

tion ﬁts in. Its main characters are:

1) It reduces the continuous stream of reality to a ﬁnite

amount of types of interpretation moments.

2) It conceptualizes hierarchically, not just by means

of collections. The hierarchy is not dictated by a

previously accepted ontology, but arises naturally in

the course of the analysis.

3) Its goal orientedness provides means to ensure ter-

minological consistency by keeping track of the pro-

cesses in which common terms ﬁgure.

Below we will present our process model for

systematic conceptualization. Since the model is

J. J. van Breemen A. and J. Sarbo J. (2008).

ON CONCEPTUALIZATION AS A SYSTEMATIC PROCESS.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - ISAS, pages 37-42

DOI: 10.5220/0001672400370042

 SciTePress

founded upon the processual semiotics of the Ameri-

can polymath Ch. S. Peirce (1839-1914) we will start

in section 2 with a treatment of some basic notions:

process (2.1) and sign and habit (2.2). This treatment

is also the occasion to argue that our way of conceptu-

alization offers organizations the opportunity to exe-

cute self control. The more technical part of the paper

starts in section 3, in which we recapitulate our model,

including its logical interpretation. An illustrative ex-

ample is presented in section 4. Finally, in section 5

we will draw conclusions.

2 HOW TO CONTROL

CONCEPTUALIZATION?

If we conﬁne our scope just to ER-modeling, it is pos-

sible to say that ER-models are used to create an accu-

rate reﬂection of the structure of part of the real world

in a database. But already if we only modestly extend

our scope and ask what did trigger the modeling of

a domain, we are forced to a more elaborate answer.

The ﬁrst thing to ask is what we do when we model,

is it capturing real world events in a reﬂection or is

it a matter of modeling our organization’s future in-

teractions with the world? Even with a narrow view

on ER-modeling the answer must be that, at the very

least, it is not only a matter of representation. For, al-

though ER developed in order to remove redundancy

and keep data consistent, if we implement a (normal-

ized) model, we affect the behavior of everybody who

is going to work for some reason with the realized in-

formation system.

With the rapid integration of information systems

across departments, the growing model of entities and

relations became hard to read. Contact with the users’

demands was restored by specifying separate busi-

ness processes, followed by a de-normalization of the

overall ER-model from the perspective of the needs

of the different roles engaged in those processes.

But how to proceed if we do not have a normal-

ized model and/or the domain to be modeled is ex-

tensive, hosting different perspectives, stated in terms

that may be vague or that lead to contradictions, in an

environment where differences of interest seriously

hamper progress? Here controlled conceptualization

may make a difference. In the remainder of this sec-

tion we will deal with our background notions as a

preparation for the presentation of the model with

which we deem controlled conceptualization feasible.

2.1 Processes and Conceptualization

The concept of process is central to our approach.

This is not primarily so because conceptualization it-

self can be regarded a process or because it is part of

an embracing process aiming at the realization of an

IS. The primary reason for taking process as a cen-

tral concept resides in the fact that all we feel, think

or do, is only felt, thought or done as part of some

process(es). From the point of view of meaning as

actualized meaning, processes are the basic units of

meaning, not terms or propositions. The latter gen-

eralize over different processes by dissociation from

any particular process. This will be clear if we realize

that disambiguation as a rule proceeds by providing

a proper context, thus embedding the ambiguous sign

in the process intended by the utterer. Stake holders

will be inclined to think about their concepts from the

perspective of the processes in which they are embed-

ded. So, explicating processes is a quite natural start-

ing point for conceptualization. Since processes ﬁg-

ure so prominently in our approach some words about

our Peirce inspired understanding of the concept of

process is called for. Hulswit succinctly summarized

the Peircean concept of process thus (Hulswit, 1999):

A process is “a continuous sequence of events

that derives its unity or internal order (distin-

guishing it from other processes) from a ﬁnal

cause, which directs the sequence to some end

state which itself may evolve.”

Several notions in this deﬁnition deserve some eluci-

dation.

A process is a continuous sequence of events:

a) Continuous because events in a process are not

separate from each other.

b) Events because there are distinguishable parts of

a process that can be articulated as facts expressed

in propositions.

Unity or internal order is derived from a ﬁnal cause:

a) The goal of the process, the desired end state, pro-

vides the Reason of the process.

b) Only what contributes to the realization of the

goal ought to be part of the model that describes the

process.

The ﬁnal cause directs the sequence to some end state:

An event starts up a goal oriented habit that governs

the process towards its terminating event.

The end state itself may evolve during the process (re-

cursively):

The terminating event needs not to be ﬁxed before-

hand in detail, it may be general or vague.

In this section we looked at processes in a very

general way, now we zoom in on the different kinds

ICEIS 2008 - International Conference on Enterprise Information Systems

of aspects that we must distinguish in all the kinds of

sign processes, including those pertinent to IS.

2.2 Signs, Habits and Conceptualization

Peirce did hold from very early on that we have no

power to think without signs (cf. CP 5.265).

But if

every thought always partakes in a sign, then the ﬂow

of thought, as far as it is connected and inﬂuential, is

representable by an argument. So, it need not wonder

us that throughout his life he kept reﬁning his thought

on signs, leaving us a tremendous corpus devoted to

the study of signs in many of its aspects. This most

certainly is not the place to trace the development of

that thought or to lay bare the ramiﬁcations of the as-

sumption. In the space at our disposal we will not

even have the possibility to deal with the bare essen-

tials of the core of his sign theory, but at the same

time we will have to present an advancement of the

basic theory which comes down to a transfer of the

concept of ‘sheet’ from his logic to his semiotics. We

start with the later and will afterwards introduce the

relevant terminology ‘on the ﬂy’.

In analogy with the ‘Sheet of Assertion’ (S

)

Pierce introduced in the context of his logic (cf. CP

4.396), we introduce a ‘Semiotic Sheet’ (S

) (van

Breemen and Sarbo, 2007). The difference between

the two is that the S

represents what is taken for

granted to be true in the Universe of Discourse by all

participants in an exchange of ideas, while the S

rep-

resents all that is believed by a group of participants or

by an individual. Let us assume that the S

has at least

two modalities. The actual S

holds what at a given

moment actually is believed, it is the state the S

in at any given moment. The potential S

contains all

beliefs that may become actualized some time.

Now, under the assumption that all thought is in

signs, lets try to describe what happens if something

demanding attention gets inscribed on a S

in the state

it is actually in. See ﬁg. 1 for a summarization of

technical terms (left) and their approximate counter-

parts in mundane language (right). Although the 9

terms suggest that they are signs on their own, this

is not the case. They are best regarded as analytical

moments that surface if we are going to analyze goal

oriented processes or reasoning.

The terms of ﬁg. 1:

The qualisign position indistinctly involves all that is

needed to reach a conclusion. Analytically it in-

volves what is offering itself for interpretation (the

effect), the actual state the S

is in and all that may

be contributed by the S

written upon in the process

A reference to (Peirce, 1931) is given by volume and

paragraph, separated by a point.

argument

qualisign

index

icon sinsign

symbol

reason

quality

connection

possibility

qualitative

existence

likeness

legisignrheme

proposition

rule

consensus

factual

actual

event

Figure 1: Sign aspects and corresponding mundane terms.

of interpretation. State and effect appear as indis-

tinct qualities.

The icon position discriminates the state and effect

according to their form, but without any interpreta-

tion. It is the single occurrence of a potential like-

ness. Looked at it from a bottom up perspective the

process can go on in many directions, none of which

is recognized at this point.

The sinsign

position indicates the co-occurrence of

the state and effect on a particular S

as an actual

event.

The index position indicates the actual connection of

all moments involved in the interpretive process, by

this it expresses the permanence of a process on a S

At the same time it expresses the latent connection

between what is actualized in this process and what

is potentially contained in this actualized S

The legisign generalizes the occurrence (sinsign) of

the instances of a form (icon). In learning this is

a bottom up process, but if the form is known the

actual event triggers a rule-like habit that classes all

the instances of that form with a type. A legisign

only surfaces by means of its instances, but as a habit

it has a real inﬂuence.

The rheme position indicates the moment the form

(icon) in its here and now actuality (sinsign) ad-

dresses the S

in order to retrieve all possible ways

in which to deal with it, to contextualize it, but only

potentially so. Creativity heavily draws upon this

moment: ﬁguring out how to possibly deal with the

input (qualitative possibility).

The symbol position indicates the moment a famil-

iar sign (legisign, involving a sinsign-icon) gets con-

nected (index) with a habit of interpretation that con-

nects the sign implied in the effect with its object.

Object here used as in “What is the object of your

remark?” The stronger this convention, the harder it

is to be creative. At this point however the habit is

still not operational.

The dicent indicates the position in which a speciﬁc

possibility is asserted for actual use in the argument

in a particular case. Out of all possible interpreta-

tions one is actually asserted (factual existence).

‘Sin’ as in ‘singular’.

ON CONCEPTUALIZATION AS A SYSTEMATIC PROCESS

The argument position indicates at least two mo-

ments. On the one hand it indicates the whole pro-

cess leading to the conclusion/response as its reason,

on the other it states the conclusion/response, ready

to enter a subsequent process.

What does this approach offer to conceptualiza-

tion? In order to establish that it is useful to introduce

a third modality of the S

, the modality of law or oper-

ational habit. What is inscribed on a S

in this modal-

ity has real inﬂuence and as such fulﬁlls the essential

function of a sign, according to Peirce:

“It appears to me that the essential function

of a sign is to render inefﬁcient relations ef-

ﬁcient, – not to set them into action, but to

establish a habit or general rule whereby they

will act on occasion” (CP 8.332).

But isn’t that the kind of sign we want to develop if

we are building an IS, at a level as much integrated as

possible? In order to realize that, we regard the differ-

ent stake holders as Semiotic Sheets that contain the

knowledge we need. When we analyze we write the

results on different sheets, for each process to be mod-

eled one. We do this by combining these sheets until

all that is needed for the processes is made explicit.

We merge and sequence the different sheets on a sheet

in which the common goals are regulative. Practical

constraints, conﬂicts of interest and ambiguities will

arise in the process. But they have a reason and can be

discussed in an understandable way, since they can be

connected to what is going on. That offers prospects

for resolution. The question we pose may be summa-

rized thus: Is it possible to develop a robust strategy

for conceptualization by taking a Peircean semiotic

account of cognitive activity as a starting point?

3 THE MODEL

In past research we suggested that Peirce’s nine types

of meaning aspects are generated in each sign recog-

nition process (Sarbo, 2007). In this section we re-

strict ourselves to a brief recapitulation of the basic

ideas.

The appearing sign (external effect) is interact-

ing with the S

occurring in a state. According to

our model, the recognition of this sign (phenomenon)

can be established through the recognition of the re-

lation between state (A) and effect (B), in the context

of knowledge about earlier interactions (¬A, ¬B) pos-

sessed by the involved S

. We assume that both state

and effect are presented as input qualities appearing

in a ‘primordial soup’ ([A B]). Sign recognition can be

modeled by a processing schema, as shown in ﬁg. 2.

(1) sorting ([A], [B]): the identiﬁcation of the two types

of qualities in the ‘primordial soup’;

(2) abstraction (A, B): the separation of the collections

of the two types of qualities;

(3) complementation ((A,¬B), (B,¬A)): the linking of

the qualities with information from the context;

(4) predication ((A,¬B)–(B,¬A)): the establishment of

a relation between the completed qualities.

(1) sorting

(3) complementation

(4) predication

(2) abstraction

[~B,~A]

[B]

[A]

(A,~B) (B,~A)

(A,~B) − (B,~A)

[A B]

Figure 2: The processing schema of sign recognition.

A logical analysis of our processing schema has

been introduced in (Sarbo, 2007). The results of this

analysis are recapitulated in ﬁg. 3. An essential ele-

ment of a logical interpretation is the abstraction of a

common meaning of the two types of input qualities,

which is the concept of a logical variable. Because

we have two types of qualities, which are indepen-

dent, the logical interpretation requires the introduc-

tion of two variables. These are denoted by A (state)

and B (effect).

~A+~B

~A*~B

A is B

A+~B,~A+B

A*~B,~A*B A*~B+~A*B

A+B A*B

A,B,~A,~B

A*B+~A*~B

Figure 3: The logical interpretation of sign recognition.

4 AN EXAMPLE

There are two dimensions along which our theory of

conceptualization can be explored. The ﬁrst is related

to its potential for merging knowledge (concepts) ob-

tained by the different stake holders representing dif-

ferent views (Lankhorst et al., 2005). This can be il-

lustrated with the speciﬁcation of a common problem

by a conceptual specialist and some domain expert(s).

Full individual speciﬁcations, each expressing a cer-

tain point of view, are merged to a single meaningful

relation, by unifying concepts of identical meaning

aspects to a single sign, through coordination. The

second way of exploration capitalizes on the potential

ICEIS 2008 - International Conference on Enterprise Information Systems

of the theory to be applied recursively, which is ben-

eﬁcial for the conceptualization of complex phenom-

ena. This section gives an illustration of such a prob-

lem. Let us emphasize that both types of exploration

are a direct consequence of the uniform representation

provided by the theory outlined in this paper.

4.1 Conceptualization of a Complex

Problem

Assume our task is the generation of a meaningful

representation of a company’s strategy. From an ab-

stract point of view such a strategy can be charac-

terized by the company’s reaction on supply and de-

mand. According to the theory of this paper, such a

reaction can be conceptualized as a relation between

supply and demand where supply can be interpreted

as the products produced by the company, which are

available for some time (state), and demand as the

suddenly appearing request dictated by the customers

(effect).

The simultaneously appearing supply and demand

(input qualities) deﬁne our company involved in a

business event. The interpretation of the two types of

qualities in themselves (individual analysis) provides

us with a deﬁnition of A and B,

and through the in-

terpretation of the relation between them (relational

analysis) we may obtain a meaningful conceptualiza-

tion of our company’s reaction to the input event. That

relation may also rely on complementary factors like

the economic perspective taken, either proﬁt maxi-

mization or purchasers satisfaction, which views in

turn are intimately related to the more primitive con-

cepts of price and selling events, that themselves can

be speciﬁed as phenomena. The next section is an at-

tempt to introduce a speciﬁcation for selling events;

an analogous treatment of the concept of price is left

to the reader.

4.2 Nested Phenomenon: Selling Event

A selling event is between a dealer and a customer.

The speciﬁcation below is given from the dealer’s per-

spective; a description from the purchaser’s point of

view is possible, but it may be less general, as the pur-

chaser’s motivation for buying or not buying a product

can be more difﬁcult to set out.

We assume that a selling event, as a phenomenon,

is deﬁned by the purchaser’s selection of certain prod-

ucts and services. The products can be interpreted as

a state, in which, the services appear as an effect. For

We refer to the status of a sign by means of the corre-

sponding logical expression.

example, dealer and customer are having a conversa-

tion while they are looking at a collection of products,

when suddenly the dealer recognizes (e.g. ‘sees’ in

a saccade), the purchaser’s selection of the services

he or she wishes to be delivered with the products.

Additionally we assume that a selling event may also

be subject to complementary factors like the different

forms of the transfer of ownership and the various ju-

dicial conditions a service may have to comply with.

The goal of this section is not the analysis of a

concrete selling event, however. Our focus is on the

interpretation of the relation between product and ser-

vice, in general. Accordingly, we will assume that the

input contains abstract qualities, which are: product

(A), service (B), form of ownership (¬A) and judi-

cial conditions (¬B). For example, ¬A may refer to

purchase, lease or rental, and ¬B to copyright, know-

how and organizational conditions (e.g. a computer

network, as a product, can be sold only as a whole,

not in parts). The representation of abstract concepts,

as qualities, is a non trivial question that we cannot

discuss in detail, because of lack of space. The ab-

stract concepts of selling are introduced as stages of

the recognition process of an abstract selling event, as

a sign (cf. ﬁg. 2).

[A]=A+B: A product (A) can be sold with a service

(B); or, a service (B), for example, a know-how, can

be available through a certain product (A).

[B]=A∗B: A product may provide a certain service

only; or, a service can be available only in combina-

tion with a certain product (product-service depen-

dency, as an actual event).

[¬A,¬B]=¬A+¬B, ¬A∗¬B: The relation between the

complementary factors of selling events like the var-

ious forms of ownership and the corresponding judi-

cial issues (¬A+¬B), and how those forms of own-

ership are regulated by law (¬A∗¬B).

A=A∗¬B, ¬A∗B: Product speciﬁcation like catalog

information about products (A∗¬B) and services

(¬A∗B), in general.

B=A∗¬B+¬A∗B: The reconciliation of purchaser’s

need for products and services, from the dealer’s

point of view. This is the concept of the habitual

way of bargaining, which is a law-like relation or

rule between products (A∗¬B) and services (¬A∗B)

that can be sold in combination. This rule includes

the possibility that products and services can be sold

independently from each other (in conformity with

the ‘or’ operation in the logical expression of B).

For example, one may buy A without B (A∗¬B) or,

B without A (¬A∗B), or both. Notice that in this

case the law-like relation identiﬁed as ‘bargaining’,

a concept that most of us are familiar with, is auto-

matically revealed by this conceptualization process

ON CONCEPTUALIZATION AS A SYSTEMATIC PROCESS

(except its name, of course). This indicates that in

cases in which we are not familiar with the concepts

they are generated by the model.

(A,¬B)=A+¬B, ¬A+B: The interrelated product and

service are actualized as constituents of a selling

event. This includes the full speciﬁcation of prod-

uct and service according to the purchaser as well

as the dealer, tacitly implicating their agreement (cf.

bargaining in context).

(B,¬A)=A∗B+¬A∗¬B: The characteristic property in-

volved in any selling event, deﬁned by a suitable

combination of product and service (A∗B) and noth-

ing else (¬A∗¬B) or, alternatively, the relation be-

tween the selected product and service (A∗B), in the

light of their ownership forms and judicial condi-

tions like the transfer of ownership through leasing

or contract (¬A∗¬B).

(A,¬B)–(B,¬A)=A is B: The meaningful description of

the input phenomenon, as a selling event.

4.3 Main Phenomenon Continued

The meaning of the nested phenomenon can be repre-

sented in the nesting phenomenon, degenerately, as

a quality. As a result, the meaningful concept of

a ‘selling event’ is reduced to complementary infor-

mation (¬B)

mediating supply and demand to their

interpretation as the company’s supply and the cus-

tomers’ demand and, ﬁnally, to a meaningful rep-

resentation of our company’s strategy. The context

of price (¬A) and selling events (¬B), is an expres-

sion of their close interaction in a market mechanism

(¬A∗¬B), but also of their possible inﬂuence as addi-

tional regulative factors, like environmental and polit-

ical demand (¬A+¬B).

The interpretation of the context along a contin-

uum marked by ‘liberal’, on the one side, and ‘state

controlled’, on the other, may provide a representa-

tion of our company (‘A is B’ as a conclusion), as a

more commercial proﬁt oriented enterprise or a more

non-proﬁt organization, respectively.

5 CONCLUSIONS

For systematic conceptualization of a domain, a pro-

cess model is needed. This has several reasons. It

minimizes unwanted ambiguity of the terms used by

consistently extracting them from the processes in

which they function. It enhances communication with

and between stake holders about their domains by the

Now ¬B, but also A, B and ¬A refer to the qualities of

the nesting phenomenon.

use of concepts in context. And it facilitates strategic

decision making in an early stage of IS building by

enabling the re-thinking of the processes that make

up any organization; after all goals can be reached in

different ways.

The approach presented in this paper is closely re-

lated to cognitive activity interpreted as a sign pro-

cess. The ﬁnite amount of types of meaningful inter-

pretation moments (cf. ﬁg. 1) facilitates the develop-

ment of a common language between stake holders,

as well as between stake holders and end-users. The

order relation between the interpretation moments as

well as between the processes increases the possibility

to control conceptualization. The goal oriented nature

of our processual model is well equipped to uncover

hidden tensions by systematically keeping check of

the relations between ends and means.

REFERENCES

Hulswit, M. (1999). From Cause to Causation : A Peircean

Perspective, volume 90 of Philosophical Studies Se-

ries. Kluwer, Dordrecht.

Lankhorst, M., van der Torre, L., Proper, H., Arbab, F.,

and Steen, M. (2005). Viewpoints and visualisa-

tion. In Lankhorst, M., editor, Enterprise architecture

at work: Modelling, Communication, a nd Analysis,

pages 147–190, Berlin (Germany). Springe-Verlag.

Liu, K. (2000). Semiotics in Information Systems Engineer-

ing. Cambridge University Press, Cambridge (UK).

Peirce, C. (1931). Collected Papers of Charles Sanders

Peirce. Harvard University Press, Cambridge.

Sarbo, J. (2007). On the logic underlying common sense.

In Cardoso, J., Cordeiro, J., and Filipe, J., editors,

ICEIS, pages 395–400, Funchal, Madeira (Portugal).

INSTICC.

van Breemen, A. and Sarbo, J. (2007). Sign processes and

the sheets of semeiosis (S

). In Liu, K., editor, 10th

International Conference on Complexity in Organisa-

tional and Technological Systems (ICOS), pages 89–

98, Shefﬁeld (UK).

ICEIS 2008 - International Conference on Enterprise Information Systems