On Advancing the Field of Organizational Diagnosis

based on Insights from Entropy

Motivating the Need for Constructional Models

Gilles Oorts, Philip Huysmans and Peter De Bruyn

University of Antwerp, Prinsstraat 13, 2000 Antwerp, Belgium

{gilles.oorts, philip.huysmans, peter.debruyn}@ua.ac.be

Keywords:

Entropy, Organizational Diagnosis, Enterprise Engineering.

Abstract:

In this paper, we explore how the ﬁeld of organizational diagnosis can beneﬁt from lessons learned from

entropy reduction in other ﬁelds. In an organizational context, entropy is related to the lack of knowledge

concerning the way of how management-level KPIs (observable system macrostate) are brought about by op-

erational elements (which are considered to be the causing microstate). Because of this lack of knowledge, the

goal and scope of projects to remedy problematic KPIs cannot be determined unambiguously. Organizational

diagnosis aims to further the insight in these decisions by providing conceptual models to ﬁnd causal explana-

tions between observations and their causes. In related ﬁelds, reduction of entropy is achieved by introducing

and analyzing structure in a system, which is described in a constructional perspective. However, we will show

in this paper that many diagnostic approaches do not support this constructional perspective adequately.

1 INTRODUCTION

Contemporary markets are characterized by volatility,

both on the demand side as a result of changing cus-

tomer preferences, as on the supply side as a result

of mergers and takeovers. Therefore, the competitive

environment of organizations is changing at a rapid

pace. Consequently, organizations need to be able

to react quickly to observed business performance is-

sues in order to satisfy customer expectations. In or-

der to be able to detect these issues, management of-

ten deﬁnes key performance indicators (KPIs) which

capture relevant scores on various criteria. When a

problematic KPI is observed, projects can be initiated

to remedy the issue, and adapt the organization to its

changing environment. However, the inﬂuencing fac-

tors of KPIs are often diverse and complex. As a re-

sult, it is not straightforward to deﬁne the concrete

scope of such projects to achieve effective improve-

ment of problematic KPI values. Various authors ar-

gue that it is indeed naive to expect that simple mea-

sures can provide insight in organizations which are

complex and variable (Sitkin et al., 1994).

The ﬁeld of organizational diagnosis attempts to

provide conceptual models to ﬁnd causal explanations

of unwanted observations (Harrison, 1994). Such un-

wanted observations can be indicated by problematic

KPIs. Without adequate understanding of the root

causes of a problem, decision makers cannot efﬁ-

ciently remedy that problem (Senge, 1990). Conse-

quently, the ﬁeld of organizational diagnosis is very

relevant in this context. However, it is still faced

with signiﬁcant challenges. First, the inherent com-

plexity of organizations makes the diagnosing activity

extremely challenging. Various authors suggest that

the search for cause and effect relations in an opera-

tional organization is very difﬁcult (Harrison, 1994;

Harry, 1988). Second, organizational diagnosis de-

pends largely on heuristics. One can expect a different

diagnosis from a novice or an experienced diagnosti-

cian. In order to better teach or develop methods for

performing organizational diagnosis, a more system-

atic approach is required.

The lack of a systematic approach and the difﬁ-

culty of handling complexity indicate the need for a

clear theoretical basis to approach these issues. A the-

oretical basis clariﬁes the concepts which are needed

to explain how complexity can be dealt with, and al-

lows to introduce prescriptive elements in a diagnosis

approach. While the selection of a certain theoretical

basis invariantly results in a focus on certain dimen-

sions, and neglects others, we believe that a relevant

theoretical basis can make signiﬁcant contributions to

an immature ﬁeld. Therefore, we explore the use of

the theoretical concept of entropy as deﬁned in the

ﬁeld of thermodynamics in order to gain more insight

138

Oorts G., Huysmans P. and De Bruyn P.

On Advancing the Field of Organizational Diagnosis based on Insights from EntropyMotivating the Need for Constructional Models.

DOI: 10.5220/0004462001380143

In Proceedings of the Second International Symposium on Business Modeling and Software Design (BMSD 2012), pages 138-143

ISBN: 978-989-8565-26-6

 2012 by SCITEPRESS – Science and Technology Publications, Lda. All rights reser ved

in the ﬁeld of organizational diagnosis. Entropy has

already been applied in a wide variety of ﬁelds. In-

sight in dealing with entropy has already matured in

those ﬁelds. Therefore, the ﬁeld of organizational di-

agnosis can progress based on lessons learned from

these ﬁelds. According to some research methodolo-

gies, such as design science, the application of proven

solutions in new research ﬁelds is the way towards

scientiﬁc progress (Hevner and Chatterjee, 2010).

This paper is structured as follows. First, we in-

troduce the ﬁeld of organizational diagnosis in Sec-

tion 2. We then explore the entropy concept and the

reduction of entropy in Section 3. In Section 4, we

apply the concept of entropy on the ﬁeld of organiza-

tional diagnosis. Finally, we summarize our conclu-

sions and the contribution of this paper in Section 5.

2 ORGANIZATIONAL

DIAGNOSIS

In organizational diagnosis consultants, managers or

researchers use conceptual models to ﬁnd causal ex-

planations of observed and unwanted effects (Alder-

fer, 2010). A diagnostician works beyond an obser-

vational role since he attempts to explain why certain

issues occur. He formulates questions (e.g., why are

ﬁve percent of the produced products defective?) and

aims to formulate adequate answers. When an en-

terprise diagnostician understands a problematic situ-

ation, a hypothesis can be formulated to explain how

an observed issue can originate. Then, evidence needs

to be gathered to conﬁrm or falsify this hypothesis.

Based on evidence, the hypothesis can be rejected or

reﬁned through an iterative process (Alderfer, 2010).

A popular approach used for diagnosing is Lean

Six Sigma (LSS). LSS applies a speciﬁc analytic

thinking pattern to support the problem solving per-

formed by the diagnostician (de Mast and Bisgaard,

2007). According to this pattern, the analytic mind

oscillates between on the one hand the theories, hy-

potheses, conjectures, ideas one has in mind (i.e., the

interpretative world) and on the other hand the obser-

vations, measurements, experimental results empiri-

cally retrieved from the real world (i.e., the factual

world) (Box and Liu, 1999). The pattern is graphi-

cally represented in Figure 1. The oscillation in this

pattern can start from any of both worlds. It could for

example start with an hypothesis one has in mind and

the gathering of facts to justify it. These discovered

facts might inﬂuence the hypothesis, which then again

needs to be justiﬁed with new facts. However, the pro-

cess can also start by observing facts from which a hy-

pothesis is built, which is then justiﬁed by new facts

Induction Induction

Deduction Deduction

Factual World

Data | Facts

Interpretative World

Theory | Hypothesis |

Conjecture| Idea | Model

Plan

DoAct

Check

Figure 1: Learning by iteration between data and models.

until a satisfactory hypothesis has been formulated.

This is called sawtooth thinking, i.e., the repeated al-

ternation of discovery and justiﬁcation in which we

develop causal explanations (de Mast and Bisgaard,

2007).

Based on this sawtooth-thinking pattern, a wide

variety of causal explanations, formulated as hypothe-

ses, can be gathered. In LSS, a so-called logic ﬁlter

is used to select the most important causal explana-

tions when faced with a business performance prob-

lem (Harry, 1988). The application of the logic ﬁlter

is organized in iterative optimization cycles. Each cy-

cle uses a speciﬁc collection of tools and techniques

to guide an applicant to the vital key correlations be-

tween inﬂuence variables and business performance

outcome variables. However, no theoretical basis is

provided to select or evaluate these tools and tech-

niques. We do not claim that this indicates that these

tools and techniques are lacking or insufﬁcient. In-

stead, we believe that a theoretical evaluation can in-

dicate more improvements in a structured way. There-

fore, the goal of this paper is to assess whether these

tools and techniques can be improved based on in-

sights from the theoretical concept of entropy.

3 THEORETICAL

FRAMEWORK: ENTROPY

In this section, we introduce entropy as a theoreti-

cal basis to interpret the current diagnosis approaches

such as LSS and to analyze how they can be im-

proved. In Section 3.1, we introduce the entropy con-

cept and its deﬁnition. In Section 3.2, we explore how

entropy is controlled in different ﬁelds. Based on this

insight, we will be able to formulate improvements

for organizational diagnosis approaches.

3.1 Deﬁning Entropy

Entropy as expressed in the second law of thermo-

dynamics is considered to be a fundamental princi-

ple. There are many versions of this law, but they

all have the same intent. Mathematical derivations

On Advancing the Field of Organizational Diagnosis based on Insights from Entropy - Motivating the Need for

Constructional Models

139

of the entropy principle start in general from a for-

mula describing the number of possible combina-

tions. In statistical thermodynamics, entropy was de-

ﬁned by Boltzmann in 1872 as the number of possi-

ble microstates corresponding to the same macrostate

(Boltzmann, 1995). The aim is to understand and

to interpret the externally observable and measurable

macroscopic properties of materials — the macrostate

— in terms of the properties of the constituent parts

— the microstate — and the interactions between

these parts. In Boltzmann’s deﬁnition, entropy is a

measure of the number of possible microstates of a

system, consistent with its macrostate. Mathemat-

ically, the entropy of a particular macrostate (S) is

equal to the Boltzmann constant (k

) times the natural

logarithm of the number of microstates corresponding

to that macrostate (lnΩ).

S = k

lnΩ (1)

In thermodynamics, examples of properties re-

lated to such a macrostate are the temperature, pres-

sure, or volume of gas in a containment. The studied

gas containment consists of a collection of molecules.

The observed values of this macrostate are brought

about by a certain arrangement of these molecules.

However, many different arrangements could result in

a certain macrostate: therefore, one cannot be sure

of the exact arrangement of molecules represented

by a single macrostate. The number of arrange-

ments which can correspond to a single macrostate is

the number of microstates referred to in the formula

above. This notion of entropy can be seen as a mea-

sure of our lack of knowledge about a system.

This deﬁnition of entropy can be further clariﬁed

by the example of a set of 100 coins, each of which

is either heads up or tails up. The macrostate is spec-

iﬁed by the total number of heads and tails, whereas

the microstate is speciﬁed by the possible conﬁgura-

tion of the facings of each individual coin. For the

macrostate of 100 heads or 100 tails, there is exactly

one possible conﬁguration, so our knowledge about

the system is complete. At the opposite extreme, the

macrostate which gives us the least knowledge about

the system consists of 50 heads and 50 tails in any

order, for which there are 10

possible microstates

(Wikipedia, 2011a). It is clear that the entropy is ex-

tremely large in the latter case because we have no

knowledge of the internals of the system.

3.2 On Controlling Entropy

A common way of dealing with entropy, is to increase

the structure or the knowledge of the internals of the

system. Consider the coin example. The entropy in

this example can be reduced when we add structure

to the studied system. Suppose we would have 10

groups of 10 coins, each with 5 heads and 5 tails, the

number of possible microstates would only be 2520

(Wikipedia, 2011a). Consequently, the entropy for

this system would be much lower. Structure can be

used to control entropy, in the sense that by allowing

less interaction between the constituing components,

a lower number of valid combinations are possible.

This leads to less uncertainty concerning the actual

microstate conﬁguration.

In complex systems, one has to consider that

structure needs to be applied to the constituent parts

of the system, not on the macrostate measurement.

In the example of the gas container, it is clear that

it would not make sense to make more detailed tem-

perature measurements. This would be an example

of a more precise macrostate measurement. Instead,

the lacking knowledge refers to the characteristics of

the individual molecules. Consequently, it is impor-

tant to be able to distinguish between the nature of the

macrostate and microstate. In systems theory, such a

distinction is made between the functional and con-

structional perspective of a system (Weinberg, 1975;

Gero and Kannengiesser, 2004). The constructional

perspective describes the composition of a system.

In a constructional perspective, the different subsys-

tems of which a system consists and their relations

(i.e., how do the different subsystems cooperate) are

described. In contrast, the functional perspective de-

scribes what a system does or what its function is, i.e.,

how it is perceived by its environment. In a functional

perspective, the input variables (i.e., what does the

system need in order to perform its functionality?),

transfer functions (i.e., what does the system do with

its input?) and output variables (i.e., what does the

system deliver after performing its functionality?) are

described. In order to reduce entropy, the structure

of a system needs to be studied from a constructional

perspective.

Models from a functional or constructional per-

spective are different in nature (Dietz, 2006). Models

created from a functional perspective are called black-

box models. These models only depict the input and

output parameters by means of an interface, describ-

ing the way how the system interacts with its envi-

ronment. Consequently, the user of the system does

not need to know any details about the inner workings

of the system. Put differently, the complexity of the

system internals is hidden in these models. Models

created from a constructional perspective are called

white-box models. These models depict the differ-

ent components of which the system consists, and

the way these components work together. Each of

Second International Symposium on Business Modeling and Software Design

140

these components can be considered to be a subsys-

tem. Consequently, each component can be regarded

as a system on its own and can therefore be described

using a functional (i.e., black-box) or constructional

(i.e., white-box) model. However, this alternation

between black-box and white-box models should be

clearly distinguished from merely adding detail to ex-

isting models. As argued by Dietz, additional de-

tail within a single perspective can be added by per-

forming functional or constructional decomposition

(Dietz, 2006). However, functional decomposition,

which elaborates on a certain model from a functional

perspective, cannot be used to obtain a constructional

model. As discussed, it is in the constructional per-

spective that the structure of a system is described.

Consequently, reducing entropy requires a construc-

tional perspective. As a result, a functional decom-

position cannot be used to reduce entropy, since the

required structure cannot be applied in this perspec-

tive.

4 APPLYING INSIGHTS FROM

ENTROPY ON

ORGANIZATIONAL

DIAGNOSIS

The concept of entropy already received attention in

management literature. Various authors applied it to

the organizational level:

• First, entropy is considered to be a measure for

waste in organizational processes. Originally, en-

tropy was a term to describe the loss of useful

energy of mechanic devices such as heat engines

when converting energy to work. Several authors

argue that waste in organizational processes can

be described similarly (Katz and Kahn, 1978).

• Second, entropy is used as a measure of uncer-

tainty with regard to a random variable in infor-

mation theory (Shannon, 1948). The so-called

Shannon entropy quantiﬁes the expected value of

a speciﬁc instance of the random variable. For

example, a coin toss of a fair coin (i.e., a coin

toss which has an exact 50% chance of result-

ing in head) has an entropy of 1 bit (Wikipedia,

2011b). If the coin is not fair, the entropy will be

lower since one can expect a certain value to oc-

cur more. In other words, the uncertainty of the

outcome has been reduced. The entropy of a coin

toss with a double-sided coin is zero.

• Third, entropy has been proposed as a measure

of industry concentration (Horowitz, 1970) and

corporate diversiﬁcation (Jacquemin and Berry,

1979; Palepu, 1985). In a concentrated indus-

try, entropy is considered to be low. The higher

the entropy, the greater the uncertaintenty will be

with which one can predict which ﬁrm will gain

the preference of a random buyer.

• Fourth, Janow has studied organizations and pro-

ductivity based on entropy (Janow, 2004). Janow

concluded that entropy offered an interesting

means to explain why organizations tend to be-

come gradually more slow in their decisionmak-

ing processes, as well as lose productivity over

time.

While the interpretation of entropy in the ﬁrst type

is related to waste, the interpretation of entropy in

the second, third and fourth types are related to un-

certainty. We will follow the latter interpretation of

entropy. For our purpose, entropy can be interpreted

as a measure of the number of microstates consistent

with a given macrostate. In an organization, a KPI

can be considered to be such a macrostate. How-

ever, when the inﬂuencing factors of this KPI are not

known, many different microstates can be relevant for

the values of this macrostate. Consequently, the en-

tropy is considered to be high.

By itself, the interpretation of KPIs as a

macrostate with high entropy does not contribute

much to the ﬁeld of organizational diagnosis. How-

ever, we can now analyze how other ﬁelds achieve

a reduction of entropy, and compare their approach

to the current practice of organizational diagnosis.

In Section 3.2, we argued that a constructional per-

spective is required to reduce entropy. Organizational

measurements such as KPIs are deﬁned in relation to

the behaviour of the organization in its environment,

and are therefore mostly described from a functional

perspective.

When insight is needed in problematic KPIs, most

approaches only propose to use functional decompo-

sition. Consider an analysis of the return on equity

(RoE) according to the strategic proﬁt model. The

RoE is deﬁned as the net income divided by the av-

erage stockholder assets. The strategic proﬁt model

proposes the DuPont formula, which breaks the RoE

down into operation efﬁciency (Net Income divided

by Sales), asset use efﬁciency (Sales divided by Total

Assets), and ﬁnancial leverage (Total Assets divided

by Average Stockholder Assets).

RoE =

NetInc.

Sales

∗

Sales

TotalAss.

∗

TotalAssets

Avg.Stckh.Ass.

However, such a decomposition does not coin-

cide with the constructional model of an organization.

On Advancing the Field of Organizational Diagnosis based on Insights from Entropy - Motivating the Need for

Constructional Models

141

Such a model will likely exist of, amongst others, the

different products. Consider a lacking product feature

as a negative impact on the sales of the organization.

In the DuPont formula, such aspects will impact both

the operation efﬁciency and asset use efﬁciency. Con-

sequently, it will be very hard to arrive at a correct

and precise analysis of the cause of the declining ROE

by using functional decomposition. Moreover, other

terms can easily be added to the DuPont formula, or

a completely different decomposition can be made.

As a result, different analysts will arrive at different

conclusions. Based on constructional models, a more

objective analysis can be made (Dietz, 2006). There-

fore, we expect the integration of constructional mod-

els in organizational diagnosis approaches. However,

different shortcomings with regard to this expectation

can be observed.

First, many causal diagrams only focus on creat-

ing ﬁner grained black-box models. Put differently,

they decompose a big black box into smaller black

boxes. However, they do not consider the relevance

of including constructional mechanisms. As a result,

these organizational diagnosis approaches limit them-

selves to functional decomposition, and exhibit simi-

lar shortcomings as described in the example above.

For example, Russo describes how Causal Loops Di-

agramming (CLD) can be used to specify correlation

relations between variables (Russo, 2008). However,

such approaches are only considered to be able to

predict the behavior of organizations, not to explain

the observed phenomenon (Craver, 2006). Moreover,

Woodward argues that such approaches may even fall

short when used for predicting behavior (Woodward,

2005): without constructional knowledge, it is not

possible to foresee the “conditions under which those

relations might change or fail to hold altogether”.

Second, certain approaches seem to propose to in-

clude constructional elements in the functional de-

composition in order to claim causality. By including

constructional elements, a direct relationship between

observed functional elements and constructional ele-

ments can be made. Craver calls such models “mech-

anism sketches” (Craver, 2006). Mechanism sketches

are incomplete models of a mechanism, which “char-

acterize some parts, activities, and features of the

mechanism’s organization, but [which have] gaps”

(Craver, 2006). With regard to this approach, several

reservations can be made.

• It has been argued that functional and construc-

tional models are different in nature (Dietz, 2006).

Consequently, different modeling constructs need

to be used, which makes a model harder to inter-

pret.

• Modeling the functional variables of an orga-

nization would already result in an enormous

amount of variables (Ettema, 2011). Adding addi-

tional variables will result in increasing complex-

ity, which makes the models harder to manage and

interpret.

• Adding constructional elements in an ad-hoc

manner fails to identify dependencies between

various constructional elements.

• Craver argues that the missing gaps in such mech-

anism sketches can function as “veil for a failure

of understanding” (Craver, 2006).

Third, approaches which explicitly incorporate a

constructive perspective, and separate it from beha-

vorial observations, do not offer any support on how

to model or select such a constructive perspective.

For example, we discussed the sawtooth thinking ap-

proach in LSS (see Figure 1). In this approach, the

behavioral measurements belong to the factual world,

while a constructional model would belong to the in-

terpretative world. However, no guidance to identify

relevant constructional elements is available: cause

and effect thinking in LSS is supposed to be per-

formed through “brainstorming” (Ettema, 2011).

This analysis shows that, in order to deal with the

presented complexity, current diagnosis approaches

(1) only consider functional decomposition, or (2) in-

clude constructional elements partially, or (3) include

explicitly constructional models, but do not provide

guidelines on how to construct them. Based on this

classiﬁcation, it can be concluded that it is useful to

develop a method, based on a current organizational

diagnosis approach, which explicitly includes a con-

crete approach for constructing constructional mod-

els (such as, for example, Enterprise Ontology (Dietz,

2006)).

5 CONCLUSIONS

In the introduction, we started by positioning two is-

sues in the ﬁeld of organizational diagnosis. We can

summarize the contributions of this paper with regard

to these issues. First, the inherent complexity of or-

ganizations makes diagnosing challenging. In regard

to this issue, this paper makes a contribution by us-

ing the concept of entropy to interpret the origin of

this complexity. Moreover, the paper shows how en-

tropy can be controlled based on insights from related

ﬁelds. We identiﬁed the presence of structure in the

constructional perspective to be primordial in control-

ling entropy. Consequently, a diagnosing approach

which attempts to address this complexity should ex-

plicitly incorporate a constructional perspective. Sec-

Second International Symposium on Business Modeling and Software Design

142

ond, no systematic approach is currently available to

perform organizational diagnosis. In order to demon-

strate this point we argued that current diagnosis ap-

proaches do not adequately incorporate the explicit

usage of constructional models. We believe that the

thorough application of engineering concepts such as

entropy in organizational research is important to fur-

ther the scientiﬁc ﬁeld described in the Enterprise En-

gineering Manifesto (Dietz, 2010). Therefore, such

a method would indeed further the ﬁeld of organiza-

tional diagnosis.

ACKNOWLEDGEMENTS

P.D.B. is supported by a Research Grant of the

Agency for Innovation by Science and Technology in

Flanders (IWT).

REFERENCES

Alderfer, C. P. (2010). The Practice of Organizational Di-

agnosis: Theory and Methods. Oxford University

Press, USA; 1 edition.

Boltzmann, L. (1995). Lectures on gas theory. Dover Pub-

lications.

Box, G. E. P. and Liu, P. Y. T. (1999). Statistics as a catalyst

to learning by scientiﬁc method. Journal of Quality

Technology, 31(1):1–15.

Craver, C. F. (2006). When mechanistic models explain.

Synthese, 153(3):355–376.

de Mast, J. and Bisgaard, S. (2007). The science in

six sigma. The American Heart Hospital Journal,

40(1):25–29.

Dietz, J. L. (2010). Enterprise engineering manifesto. On-

line available at: http://www.ciaonetwork.org/

publications/EEManifesto.pdf.

Dietz, J. L. G. (2006). Enterprise Ontology: Theory and

Methodology. Springer.

Ettema, R. (2011). Diagnosing the enterprise, towards a true

causal claim. In Proceedings of the 2011 CIAO DC.

Gero, J. S. and Kannengiesser, U. (2004). The situ-

ated function-behaviour-structure framework. Design

Studies, 25(4):373–391.

Harrison, M. I. (1994). Diagnosing Organizations: Meth-

ods, Models, and Processes. Sage.

Harry, M. (1988). The nature of six sigma quality. Motorola

University Press.

Hevner, A. and Chatterjee, S. (2010). Design Research in

Information Systems: Theory and Practice, volume 22

of Integrated Series in Information Systems. Springer.

Horowitz, I. (1970). Employment concentration in the com-

mon market: An entropy approach. Journal of the

Royal Statistical Society, 133(3):463–479.

Jacquemin, A. P. and Berry, C. H. (1979). Entropy measure

of diversiﬁcation and corporate growth. The Journal

of Industrial Economics, 27(4):359–369.

Janow, R. (2004). Shannon entropy and productivity: Why

big organizations can seem stupid. Journal of the

Washington Academy of Sciences, 90(1):39–50.

Katz, D. and Kahn, R. (1978). The social psychology of

organizations. Wiley.

Palepu, K. (1985). Diversiﬁcation strategy, proﬁt perfor-

mance and the entropy measure. Strategic Manage-

ment Journal, 6(3):pp. 239–255.

Russo, F. (2008). Causality and Causal Modelling in the

Social Sciences: Measuring Variations (Methodos Se-

ries). Springer.

Senge, P. M. (1990). The Fifth Discipline: The Art and

Practice of the Learning Organization. Doubleday.

Shannon, C. (1948). A mathematical theory of communi-

cation. Bell System Technical Journal, 27:379–423.

Sitkin, S. B., Sutcliffe, K. M., and Schroeder, R. G. (1994).

Distinguishing control from learning in total qual-

ity management: A contingency perspective. The

Academy of Management Review, 19(3):537–564.

Weinberg, G. M. (1975). An Introduction to General Sys-

tems Thinking. Wiley-Interscience.

Wikipedia (2011a). Entropy.

Wikipedia (2011b). Entropy (information theory).

Woodward, J. (2005). Making Things Happen: A Theory of

Causal Explanation. Oxford University Press, USA.

On Advancing the Field of Organizational Diagnosis based on Insights from Entropy - Motivating the Need for

Constructional Models

143