This report looks to understand the fundamental
concepts of modelling and how it can then be used in a business. This was done with a series of rational arguments from a set of definitions and axioms
through a set of propositions and finally a conclusion. It finds that a model
must necessarily be created from information, processes, systems and
relationships. That none of those concepts could be removed from a model and
each of them must be made specific in order to deliver in an environment with finite
resources. The opening argument looks at if a model can be developed using process
alone. This is invalidated by the necessary join between process and
information within physics and linguistics. A second proposition looking at
only information to model a business is invalidated for the same reasons. Next
a view of creating a model with only information & process is proposed but
this cannot work as it creates models with no concept of agency. The concept of a system is
added alongside information and process but there is no means of joining those
concepts together and therefore a model that offers completeness in its
description of something cannot be created. Finally, a proposition that says a model
must contain information, process, systems and relationships is proposed and no
compelling counterarguments to this proposition have been found. Building upon
that there is the view that because each of those concepts necessarily has relationships
that there are interfaces. Those interfaces have the potential to be affected
by noise and therefore must be controlled. As each modelling language has to
make each of these concepts specific in order for efficient delivery it will
necessarily bias itself towards certain technological architectures.
Practitioners and managers alike should be aware of each of the true
propositions when assessing an existing modelling capability or when looking to
develop a modelling capability.
We all live in a world where we must work and communicate
with a vast array of people and an ever-increasing number of electronic
systems. Each one of those people and systems have their own way of
understanding the world which when grouped together are forced to develop
common forms of communication often termed languages. Beyond the natural
languages such as English, Russian and Chinese to name but a few there are also
languages developed artificially to support the world of business and
technology. These languages used by professionals such as business analysts,
enterprise architects, software developers and managers, describe the world in
the past present and future. They do not however all speak the same language
and as such, there is the potential for miscommunication.
This report looks to understand the fundamental concepts
underpinning the languages of each of those professions use to work. To be able
to convey that understanding to give each of them a more sophisticated means of
communicating, as well as offering guidance on how managers can understand
their businesses modelling capabilities and conduct the necessary analysis to
improve their business’s performance.
The report covers basic background research done into how
different models try to describe the world as well as efforts to draw together
meta-models of those as well as the steps taken to prepare this report. It uses
three specific modelling languages used by different professions to ensure the
concepts exist in multiple languages each with a different purpose. The
selected languages are ArchiMate which is used by enterprise architects, Business
Process Modelling Notation which is used by managers & business analysts
and Uniform Modelling Language is used primarily by software engineers.
The core structure of the report is based on Ethics by
Spinoza. This looks to give the definitions of each of the fundamental concepts
that are unique and independent. Then a set of axioms, accepted truths, that
all models must adhere to. Then a set of propositions that look to link the
concepts together and align them with the axioms put forth.
An evaluation into the limitations of the work will then be
performed. Followed by a conclusion wraps up each of the propositions caveat it
against any noted issues from evaluation.
In order to devise the fundamental concepts of modelling an approach
was taken to understand a few different business domains. To select a modelling
language from each of the domains, with the chosen domains being process
modelling, architecture and software engineering.
Then to draw out the concepts conveyed in each and to
generalise them. Those general concepts would then need to be validated and
defined using knowledge beyond each of the original modelling languages. To do
this each concept will be looked at from a colloquial, philosophical, and
scientific viewpoint. In order to gain a viewpoint within the time available,
high-level summary articles will be sought, viewpoints could be refined in the
future with more detailed research.
The structure of the report was selected before the key
concepts were gathered. That structure is mentioned in the introduction taken
from a philosophical work that aims to understand fundamental concepts of the
universe which parallels the arguments made within this report but at a
different level of abstraction. With the structure already selected and based
around a logical series of propositions, the debate behind each of those
propositions is what guided the next stage of research.
Any questions raised about the definition during the
investigation into each proposition would mean going back to the start,
reworking the definition and then revisiting each of the propositions. This
process was repeated until none of the propositions were invalidated.
A description of the selected modelling languages and their
purpose.
ArchiMate is a modelling technique ("language")
for describing enterprise architectures. It presents a clear set of concepts
within and relationships between architecture domains and offers a simple and
uniform structure for describing the contents of these domains. [1]
ArchiMate is primarily used by an enterprise, business and
solutions architects trained within the open group’s framework and has tools
that support its use alongside TOGAF (an architectural framework).
“The Business Process Modelling Notation (BPMN) is a
graphical notation that depicts the steps in a business process. BPMN depicts
the end to end flow of a business process. The notation has been specifically
designed to coordinate the sequence of processes and the messages that flow
between different process participants in a related set of activities.
BPMN is targeted at a high level for business users and at a
lower level for process implementers. The business users should be able to
easily read and understand a BPMN business process diagram. The process
implementer should be able to adorn a business process diagram with further
detail in order to represent the process in a physical implementation. BPMN is
targeted at users, vendors and service providers that need to communicate
business processes in a standard manner.” [2]
“The OMG's Unified Modelling Language™ (UML®) helps you
specify, visualize, and document models of software systems, including their
structure and design, in a way that meets all of these requirements. (You can
use UML for business modelling and modelling of other non-software systems
too.)” [4]
The Uniform Modelling Language is primarily designed for the
development of software as stated above in an object-orientated fashion. This
means it focuses on how it describes objects, called classes, that are capable
of executing functions and storing data. This will be primarily used by
software engineers in developing coded applications although UML can be
repurposed as a business modelling language. It does not prescribe to any
particular methodology to develop the models, those have to be developed
separately.
Definition: A statement of the exact meaning of a word.
Each of the definitions below are aiming are for unique and
independent concepts found in current business modelling languages and validated
with colloquial, philosophical and scientific viewpoints. The detail for each
definition is aimed at lending understanding to the reader as well as some
general properties about the concepts. Those described properties will then be used
to understand how each of those independent concepts interrelate in a model.
There is great difficulty in understanding what information
is because of the plethora of definitions and the fact a lot of them are bound
in specific contexts for example Data being defined as ‘quantities, characters
or symbols on which operations are performed by a computer’. The first section
is fine however ‘performed by a computer’ adds a constraint that is plainly
wrong. It is perfectly possible for a human to perform the same operations on
data that a computer does.
Data: facts and statistics collected together for reference
or analysis or the quantities, characters or symbols in which operations are
performed by a computer, which may be stored and transmitted in the form of
electronic signals and recorded on magnetic, optical, or mechanical recording
media [2]
This gets more confusing when the Definition of information
is thrown in as a synonym of data and thus allows the two to be used
interchangeably.
Information: facts provided or learned about something or
someone or data is processed, stored, or transmitted by a computer [3]
What each definition agrees upon is that information is a
thing, a noun in language terminology. That information has the ability to
describe something and that description can be made up of a variety of characters,
symbols or numbers. Those descriptions fit into one of five categories: natural
language, optimal codes, numbers, physics and logic. [4]
Each of those types of information have their own areas of
research and understanding however they generalise into two types, Information
that has semantic meaning and information without semantic meaning. Information
without semantic meaning can be measured in bits. Where a bit has a state that
is either true or false. Each of the five categories of information can be
written using a collection of bits. [5]
semantic information cannot be measured however in order to be
represented and communicated it has to be put into a form that is communicable.
That communicable form is always one of the non-semantic information types.
This leaves us with information as a thing of the fixed
state that can convey a description. That there may be many interpretations of
a set of information but in that information has to be represented by a fixed number
of bits, structured using a natural language, optimal codes, numbers, physics
or logic.
There remains a final and complex point. Information being
something that can describe means that the description of information is itself
information. Although non-semantic information can break down to the
fundamental level of bits, it is possible for us to describe what bits are
using bits, meaning we can continue within that particular loop forever.
Understanding how exactly this can happen is unimportant but understanding that
information is capable of self-reference is.
Information: a thing of state that can describe, often
represented using symbols, numbers and code in a language and measured in bits.
ArchiMate
“Structure elements are the strategic element resource, and
structural elements, which can be subdivided into active structure elements and
passive structure elements” [7]
Information within ArchiMate that relates closest to our
description of Information is the noun based elements within the ArchiMate
model. There are two types of structure active and passive as you’ve seen in
our investigation into information so far it is prescribed no agency of its own
therefore it is the passive structure element that aligns with our view of
information.
“A passive structure element is a structural element that
cannot perform behaviour. Active structure elements can perform behaviour on passive
structure elements. Passive structure elements are often information or data
objects, but they can also represent physical objects.”
ArchiMate may differentiate between physical, logical or
conceptual objects within its model but the general definition lines up with
our view that information is the stateful concept of description, and has no
agency of its own.
Business Process Modelling Notation
“Data Objects provide information about what activities
require to be performed and/or what they produce, data objects can represent a
singular object or a collection of objects.” [8]
BPMN suffers from the information/data confusion mentioned
earlier though at this point it places information as a component of data. This
does not invalidate our views as we have already pointed out that is possible
for information to describe information especially when the information has a semantic
meaning which is not attached to the limitations of the physical world.
Uniform Modelling Language
“A classifier describes a set of objects. An object is an
individual state and relationship to other objects. The state of an object
identifies the values for that object of properties of the classifier of the
object.” [9]
In UML classifiers are a describable thing. As UML was
created to allow the models to create software, a classifier acts as a description
of a possible set of objects. Classes can contain functions, our view of
information within UML would be for classes without functions. Within a model,
these would usually be represented by an entity-relationship diagram. This
would then hold to the view looked at earlier and seen in the other modelling
languages that information is that which can be described and remains in a
fixed state unless acted upon. There is also inbuilt in the definition the self-referencing
required the concept of information. In this case, a classifier describes an
object and an object is a group of properties/attributes which could be
represented within a computer by bits.
Process suffers from a similar fate to data and information as
it can be expressed with a multitude of words. Such as procedure, activity,
task and operation which are but a few of the other possible words that fit
within the same definition.
Process: series of actions steps taken in order to achieve a
particular end [10]
Each of these words are verbs that convey the concept of
change, each one breaks down into a subset of instructions that take one state
of the world and turn it into another. We can, therefore, take two things, one process
is something that changes the state of things, two it can be broken up into a
series of smaller changes which are often seen as instructions or action steps
or tasks. That breakup should be seen as self-reference, the same concept of change
can be decomposed. This avoids defining change as a hierarchy of different
words that all have the same meaning.
Process unlike information has less exact philosophical or
mathematical constructs or at least vaguer more rational approaches. Process is
often given different names usually the names of the discipline themselves
which are often seen as the process or undertaking of discovering something or
expanding our knowledge about the world. For example, business is the process
making exchanging goods and services for money. [11]
Within science, General Relativity, and Quantum theories
both contain the concept of change, generally measured using time, of course,
change within those mathematical models is not related to just time but time is
the process of change that we experience and can most easily relate to.
Within each of the models, the concept of change was looked
for and separated where possible from the thing it changes or the thing that
performs the change. Each of these definitions should ideally be able to be
decomposed into smaller concepts of change.
The colloquial, scientific
and modelling languages use of process all align fairly closely as a change to
something that can be broken down or compounded into larger or smaller chunks
of change. Where change is something done to a state to move it to another
state.
Process: Means of changing a state.
ArchiMate
“This generic metamodel fragment consists of two main types
of elements: structure (‘nouns’) and behaviour elements (‘verbs’)” [7]
The highest level concept of change within ArchiMate is what
it terms behaviour elements. These like our analysis align with verb like
concepts that are considered to bring about change.
“Behaviour Element: element that represent a dynamic aspect
of an enterprise.” [13]
This is expanded further detailing behaviour elements as the
dynamic aspect of an enterprise, the piece of the business that is capable of
changing. How that change is broken up is not mentioned.
“Process: a series of behaviours that achieve a specific
outcome.” [14]
The lowest level definition of change is process, which
aligns with our view of change including self-reference. This is shown by the
fact the process is broken up into a series of behaviours that achieve a
specific outcome. Which aligns almost exactly to our view of the concept change
from one state to another that can be broken down into smaller concepts of
change.
Business Process Model
Notation
“A process depicts the interactions between two or more
business entities. These interactions are defined as a sequence of activities.”
[15]
Once again, we can see a concept of change performed to a stateful
thing and that this change can be broken down into smaller concepts of change.
“An activity is a generic term for work that a company
performs.” [16]
“A Task is an atomic activity that is included within a
process. A task is used when the work in the process is not broken down to a finer
level.” [16]
The same effect can be seen to go down another two levels
from the definition of process through activity all the way to task, that considered
to be atomic although tasks are defined as a set of steps. The fact that any
business model has to avoid trying to describe things infinitely is an
unnecessarily confusing part of many models.
Uniform Modelling Language
“Behaviours. A behaviour describes a set of possible
executions. An execution is a performance of a set of actions (potentially over
some period of time) that may generate and respond to occurrences of events,
including accessing and changing the state of objects” [9]
The highest level within UML ‘s behaviours is the concept of
change that acts upon a state, in this case objects, and can be broken down
into smaller units of change described as executions. Multiple levels of change
is shown further in the way that executions can be broken down into actions.
System, as a concept, unlike process and information doesn’t
suffer from a multitude of terms but it does suffer from a general vagueness.
There are a huge number of fields that happily use the word system but very few
define exactly what they mean by it. Generally, the meaning of system within
any particular piece of work is considered to be implicit.
System: a set of things working together as parts of a
mechanism or an interconnecting network; a complex whole [16]
Given that no scientific or philosophical consensus exists
regarding what a system is. One of the many fields that define systems is
Cybernetics, which builds on the general theory of communication and looks to
understand interaction and control within animals and machines. The view in
cybernetics on systems links to the colloquial understanding, a set of things
that form a complex whole. As a multidisciplinary field Cybernetics looks to
see if issues in both machines and animals can be compared. The central idea is
that both animals and machines are the same thing and that the actions or
mechanics of both things are therefore comparable.
“Computing machines are essentially machines for recording
numbers, operating with numbers, and giving the results in numerical form.” [17]
The quote above places machines as things that operate on
information that is given to them and that can provide a response back out
again. if we take animals and machines as the same and say that they are
capable of taking in information, changing information, storing information and
outputting information. All of these actions on information also indicate that
systems are capable of change and therefore of performing or being a part of
processes. Systems are characterised as black and white boxes.
”Blackbox and white box are convenient and figurative
expressions not very well determined usage. To understand by black box a piece
of apparatus which performs a definite operation on the present and past of the
input potential, for which we do not necessarily have any information of the
structure by which this operation is performed. On the other hand, white box
will be a similar network in which we built in the relation between input and
output potentials in accordance with a definite structural plan for securing a
previously determined input-output relation.” [17]
The explanation above again gives the view of a thing that
can take in information perform processes and output information. This makes it
a grouping of both, in the white box the internal processes and information can
be understood and in the black box they remain a mystery but the interfaces can
be understood. This can align with the colloquial definition of system a
collection of things that we can refer to as a complete whole that perform
actions or at least take part in processes.
System: a group of descriptive states capable of internal and
external change that can be prescribed with agency
ArchiMate
“Active structure elements are the subjects that can perform
behaviour. These can be subdivided into internal active structure elements;
i.e., the business actors, application components, nodes, etc., that realize
this behaviour, and external active structure elements expose this behaviour to
the environment. An interface provides an external view on the service provider
and hides its internal structure.” [18]
ArchiMate describes a system in terms of active structure
elements that are capable of performing behaviour, they also can act as an
interface a point of transferring information from one place to another. It
differentiates between active structure elements that are considered to be
inside the model and those that exist beyond the model or indeed the business.
Business Process Modelling
Notation
“Participant: A business entity (e.g., a company, company
division, or a customer) or a business role (e.g., a buyer or a seller) that
controls or is responsible for a business process. If Pools are used, then a
Participant would be associated with one Pool. In a Collaboration, Participants
are informally known as “Pools”.” [19]
The concept system within business process modelling
notation realised by a participant. Participants always interact with one
another or information via process or messages.
Uniform Modelling Language
“A model is always a model of something. The thing being
modelled can generally be considered a system within some domain of discourse.
The model then makes some statements of the interest about that system,
abstracting from all the details of the system that could possibly be
described, from a certain point of view and for a certain purpose. For an
existing system, the model may represent an analysis of the properties and
behaviour of the system. For a planned system, the model may represent a
specification of how the system being modelled is to be constructed and
behave.” [9]
As seen previously classifier is within UML are capable not
only of storing information but also performing functions. This lines up with
our definition of a system being able to store information also be an agent of
change. There is also an alignment with UML’s definition of a model as
something that store state and is capable of internal and external change.
Each view of relationships including the colloquial
generally looks at how a collection of similar or different concepts join our
link together.
“Relationship: the way in which two or more people or things
are connected, or the state of being connected” [20]
A very formal definition of relationships is to use logic,
that makes statements about the relationships between concepts. The mechanics of
logic form the underpinnings that allow us to understand the grammars of
natural languages, mathematics and physics.
“At their most basic, logic is the study of consequence, and
information is a commodity. Given this, the interrelationship between logic and
information will centre on the informational consequences of logical actions or
operations conceived broadly” [20]
Now unlike process, logical action is not one of state change
but one that asserts a link between two pieces of information to be true or
not. For example, we could have a relationship logic of ‘A is in B’ which when
filled out could be the ‘page is in the book’. Stating that a piece of
information links to another piece of information and not at any change or
process has been performed on either piece of information.
Relationship: A connection or link between concepts
ArchiMate
“relationships, each of which can connect a predefined set
of source and target concepts (in most cases elements, but in a few cases also
other relationships). Many of these relationships are ‘overloaded’; i.e., their
exact meaning differs depending on the source and destination concepts that
they connect.” [19]
ArchiMate uses a wide range of relationships which as
defined above created using a source and target for any type relationship. When
compared to a definition we see both align as a link between concepts.
Business Process Modelling
Notation
BPMN again makes use of relationships and has different
types depending on the relationship between information and flow objects or
flow objects with other flow objects. Flow objects, in this case, being a
process, activity or task.
Figure 1
Example Process flow with Associations, Sequence flows and Message Flows. [22]
Primarily sequence flows, the solid black lines with arrows
on the end go-between each activity boxes and convey the order in which those
activities should take place. Associations, the dotted lines between each
activity. The data objects are visualised as letters. [22]
Uniform Modelling Language
“Association is a relationship between classifiers which is
used to show that instances of classifiers could be either linked to each other
or combined logically or physically into some aggregation.” [9]
The same is seen in UML where associations are used to link
classifiers to one another.
A model will be a collection of concepts where the rules by
which those concepts relate to one another are defined by a language. This can
be seen across topics mentioned before such as logic, mathematics, physics and
natural languages.
“Model: A simplified description, especially a mathematical
one, of a system or process, to assist calculations and predictions.” [20]
The specific science of philosophical view of models
generally align directly with the colloquial definition from the dictionary
given above. Model will serve as a simplified description of something. Within
this particular piece of work those somethings will be concepts as is looked at
within the definitions chapter.
Model: a group of concepts that form a description of something
as defined by a language.
ArchiMate
“A collection of concepts in the context of the ArchiMate
language structure.” [20]
This definition is similar to the one given in ArchiMate,
where we can see that a collection of concepts grouped together using the rules
of the language can describe a particular context.
Business Process Modelling
Notation
There is no specific definition of a model is given within
the BPMN specifications.
Uniform Modelling Language
“A model is always a model of something. The thing being
modelled can generally be considered a system within some domain of discourse.
The model then makes some statements of the interest about that system,
abstracting from all the details of the system that could possibly be
described, from a certain point of view and for a certain purpose. For an
existing system, the model may represent an analysis of the properties and
behaviour of the system. For a planned system, the model may represent a
specification of how the system being modelled is to be constructed and
behave.” [9]
Less concise than the ArchiMate definition but it speaks
along similar lines. That a model is describing something. That something is
detailed in a collection of concepts that have a particular purpose or
viewpoint. It should be noted that the particular purpose or viewpoint has not
been added to our own definition. This point is considered as an additional
property that may or may not be necessary. The answer to which is looked at
later in this document.
Axiom: A statement that is taken to be true, to serve as a
premise or starting point.
As the definition states above each of these are to give a
starting point to the propositions. They said the high-level limits by which
each of the conclusions of the propositions must have dealt with.
This axiom is hopefully the easiest to argue for. Views and
understanding within philosophy and science and business should move with one
another. An extreme example of this is the following: Currently, it is not
compelling to build into any business case a means of persuading Poseidon for
less storms and a larger yield on fish. More recently as the rise of technology
perform roles used to be work by people, in a competitive world it does not pay
to stick to tradition.
“The despotism of custom is everywhere the standing
hindrance to human advancement, being in unceasing antagonism to that
disposition to aim at something better than the customary” [22]
As per above by John Stuart Mill if we no longer have
reasons for holding of you in that view has become custom, we should do
everything we can to understand the views we have and if we no longer have
reasons to understand those views then we should look to replace them. Or as put
by philosopher Bob Dylan “The times they are a changin“.
The definitions given for information, process, systems and
relationships have all been done to be independent of one another and based on
philosophical, scientific and colloquial knowledge. As concepts , within our
imagination, they are able to continuously break down and compound with no
limit. Just like the scene with the information that can describe information
and the process that can be broken down into processes there also seems no
reason why our model cannot contain each of the concepts discussed breaking
down infinitely and interacting infinitely between themselves.
This does not mean that this is not a problem or at least
something to consider when looking through and arguing all is fundamental to a
model but it seems like something that is necessary to point out in order to
propose means of solving it.
Businesses are constrained by the amount of money, people
and resources available to them. This means that any endeavour to model
businesses will be subject to the same limitations.
Any pragmatic view on developing a model must necessarily
take heed of the fact that there is not an infinite amount of time people money
and resource to perform that endeavour. Modelling languages and practitioners
are therefore looking at the balance between the infinity of concepts that
encompass the best possible understanding of the world for the least amount of
money time and resource.
Proposition: A statement or assertion that expresses a
judgement or opinion.
Each of the propositions below is targeted at leading the
reader to a conclusion via a series of propositions about the fundamental
components of a model. Each proposition is an argument that may reference others
in order to build upon a growing theme.
Could a model of a business be constructed from process
alone? As we have seen our definition process is “Means of changing a state.”. A
cursory glance at physics will point out to us that change alone is not
possible.
“General relativity tells us that what we call space is just
another feature of the gravitational field of the universe, so space and
space-time can and do not exist apart from the matter and energy that creates
the gravitational field.” [22]
As we noted with our first axiom models should align and
take understanding from the best view we have of the world at the time. It
would seem that not only can our latest models of the universe from physics that
process does not exist alone furthermore it is intertwined with physical
substance in the form of space-time.
Should this example not be compelling or considered too
high-level an example closer to how we operate on a daily basis comes from natural
language as documented by linguistics. Within linguistics, our natural
languages are categorised according to their basic word order which is
constructed from (subjects, verbs and objects).[22]
Within the field of linguistic typology, there have been
searches for asymmetric languages that do not follow or miss a particular part
of the SVO construct, this however has only ever been seen with certain types
of object/subject missing but not every kind of object/subject. [23] It would seem that
the concepts of change cannot exist independently from concepts of state.
The model most likely to exist on change alone is BPMN which
despite having a limited ability to model data objects, messages and
participants it still keeps these constructs within its model. Both ArchiMate
and UML have concepts other than just process within their language. All this
invalidates the initial proposition, process alone cannot model a business.
Process alone cannot model a business.
Given the model could not be constructed from process alone.
Is it impossible to construct a model from information alone where information
is “concept of a thing of state that can describe, often represented using
symbols, numbers and code in a language and measured in bits”?
The same arguments can be applied as shown above, because as
it was pointed out that the concept of change in time is inextricably linked to
space. Given that space is physical and information is used to describe physics
then the fact that within those models of physics, state and change cannot be
separated the same would apply to information’s ability to separate itself from
process. [25]
The view of linguistics is similar, there are examples of
languages without certain types of verb but they do not lack every type of
verb. This time pointing out that information or things of state are also stuck
with process. [27]
Only one of the modelling languages allows for a model to be
constructed of only information, that model being UML. An example of which can
be seen below.
Figure 2
Boxes = Entity, Lines = Relationships, Colours = Domain
There are two major issues presented to a proposition within
the diagram above. The first is that despite this being a collection of
entities we still require another concept to explain it. In this case, the
concept being a relationship, although as we have already said information can
be self-referencing the modelling language itself requires a relationship to
demonstrate how each of these objects relate. The second issue is that these
diagrams are used most commonly to structure information within a database for
example SQL Server. Within SQL Server that data model is interacted with using
a query language that allows you to create, read, update and delete from each
of those entities.
From a technical point of view then it is important to
understand the processes that interact with the information modelled above are
still required and from a business point of view it would also be important to
know who or what owns and manages the information modelled. Therefore,
information alone cannot model of business and proposition once again is
incorrect.
Information alone cannot model a business.
Given that Prop1 (Process alone cannot model of business)
and Prop2 (Information alone cannot Model a Business) then perhaps the two
together can. Leading to propose that process and information can model a
business.
Within Cybernetics based on the general theory of
communication, there are objects that perform processes but also objects that process
& store information. This encapsulates the description given before of
Black boxes and white boxes. Now although process and information can describe
each of those boxes thus far we have no concept of grouping the two together. Furthermore,
our definitions of processing information do not allow for agency or the idea
that a thing is capable of acting of its own volition. Process does not cover
this because it is merely the concept change and going from state A state B
does not require any understanding of what performs a process. Information does
not cover this because it is a fixed state unless acted upon and therefore
necessarily cannot act upon other states via processes. [16]
In Linguistic, verbs are analogous to process and objects
are analogous to information then there remains subject. Although generally, it
is possible for the subject to be a verb or object. In the case of the subject
playing a semantic role it is considered that the subject has agency which is a
property beyond process and information. [28]
ArchiMate has a concept of active structure which is
considered to be things that perform processes but also store information. They
generally considered as interfaces which do not fit the definition of either a
process or information that are the combination of both. In addition, there are
also concepts of behaviour which are not merely just process, these include
services and functions which are capable of performing processes and are also
capable of storing and changing information. [17]
UML’s class is not a process or information even at a
potential level, its capable of capturing both
The primary concept within UML is the classifier which does
not symbolise process or information it constitutes both of them. This means
that the proposition of information and process alone being enough to create a
model is false. [8]
The other examples in order then BPNM has within it
participants which are neither data objects or processes but are capable of
performing processes bus prescribing them agency which again is not covered by
process or information. Therefore, process and information alone cannot model
of business.
Process & Information alone cannot model a business
Given proposition three (Process & Information alone
cannot model a business) the next proposition should support the additional
concerns raised. The concept required is that of a system which we defined
earlier as “a group of descriptive states capable of internal and external
change that can be prescribed with agency”. The addition of system then means
that we can say what is using information, what is performing processes. Where
those systems could be as stated by Norbert Weiner animal or machine. [16]
There is however an issue with what we’ve just described we
have process, we have information, we have systems, we said that somehow it is
possible for information to be changed by process, we said it is possible to
group information and processes into systems, and finally, it is possible for
systems not only be composed of processes and information but to beyond their
boundary play a part in processes and interact with external information.
Each of our definitions shown thus far are a concept of
change, of state and of a group. None of which allows us to say anything about
how particular instances of change, state and group could relate to one
another. In our example so far both physics, Cybernetics and linguistics have
not only the concepts we discussed but a means of joining them together. In the
case of physics and Cybernetics the means of joining those concepts together is
played by a series of logical and mathematical statements. Linguistics, on the
other hand, has a grammar that denotes the possible relationships between each
of the subject-verb-object concepts. [16] [28]
What a model would require in order to join together each of
those concepts is a concept of togetherness, a link if you will. This was also
covered as one of our definitions, our concept of relationship.
Every single one of three modelling languages, Archimate,
BPMN & UML make use of relationships to join process, information and
systems. An example from automate shown below which shows how an event leads to
a series of processes being started how those processes at least some of them
interact with information. There are other examples within every single one of
the modelling languages could be sought out the point should be evident enough
by now.
Figure 3
Access path example from ArchiMate. How could this be done without
relationships? [29]
Concluding this particular proposition we can now say that a
business cannot be modelled with processes, information and systems alone.
A Business cannot be modelled with Processes, Information
& Systems alone.
Propositions 1 to 4
showed how each one of these unique and independent concepts cannot alone describe
everything about the model and that they must necessarily all exist for a model
to give a complete view of a business.
The first axiom
covered that any model must move with the knowledge of the times, this been
covered in our initial definitions supported with colloquial, philosophical and
scientific examples where each of these concepts exist. Not only that in each
of the propositions the falsification of each was achieved with examples from
philosophy or science.
This will give us
an overall model of models something like the diagram below.
Figure 4
Processes, Information, Systems & Relationships can model a Business.
ArchiMate, BPMN & UML can each have their concepts
translated into either process, information, systems or relationships with no
exceptions. They will, however, break apart their model as noted when each of
them was introduced for specific purposes and to ensure they can offer value
rather than being purely an exercise in reasoning.
The second axiom (Concepts are infinite) we noted that
concepts without the bounds of reality would be infinite and that any
reasonable model should be able to encapsulate infinity. The above diagram
indicates how each of the concepts can exist and how they relate to one another
and that because each can relate to one another and themselves then there is
infinity for each concept and the model in general. For a business, however,
this represents a problem as noted in axiom three (Money, Time & Resource
are finite) we have finite money time and resource with which we can model. So
the proposition that a business can be modelled process, information, systems
and relationships are logically correct there is still some further
investigation as to how to deal with the infinities presented by the
proposition.
Processes, Information, Systems & Relationships can model
a Business
In order to take our conclusion from proposition five (Processes,
Information, Systems & Relationships can model a Business) and ensure that
the conceptual infinity noted in axiom two (Concepts are infinite) can be dealt
with in such a way so that it can coexist with axiom three (Money, Time &
Resource are finite)) it will be necessary for us to understand the bounds of
any particular piece of modelling work and if necessary create a more detailed
modelling language from the abstraction from prop5 to ensure that the efforts
of modelling provide value to a business.
This can be seen in your examples where specific key trends
can be seen to limit the amount of time required to model and also to place
complexity into defined boxes. Active structures as noted before used in
ArchiMate act as interfaces between the business being modelled and the
external world. This forms the first key boundary that each model covers which
is between what is being modelled and what is beyond that model. However,
depending on the scope of a modelling project it is also possible for internal
interfaces between sections of any model to exist. This is particularly
prescient when looking at the image below.
Figure 5
Aspects and layers as defined in ArchiMate [30]
The way in which each of the concepts is made more specific
or applied to a particular area of the business will depend on the purpose of
modelling taking place. The aspects and layers of ArchiMate enable the
development of an enterprise, business, application and technology
architecture. In doing so the purpose of the model to give a high-level view of
each of the areas as well as a means of linking all those areas back together
into a single picture is achieved. There are other ways of cutting up the
concepts for specific purposes which can be seen in the image below.
Figure 6
UML Semantic Area. [31]
The purpose of our UML models to looking at a predefined
area then offers a series of likely models that will be helpful in capturing or
showing a particular view of the business based on the purpose that the
exercise has initially been given. It would be possible to use each of these UML
semantic areas to model each of the aspects and layers of an ArchiMate model
rather than the language ArchiMate prescribes.
This offers the ability to understand a set of predefined
deliverables they set purpose that is achievable within the finite money, time
and resources available to a business and avoid the potentially infinite amount
required if a complete model is asked for. It should however not be forgotten
that there are core concepts behind a model that should always be considered
and known even when those contexts and boundaries are set.
Models must set Context & Boundaries
The first axiom points out that any business modelling and
pay heed to the knowledge and understanding we have of the world at the time.
Given that as stated in proposition five (Processes, Information, Systems &
Relationships can model a Business) the need for process, information, systems
to relate then there are going to-be interfaces. Each of these interfaces will
be affected by a concept called noise as defined by Claud E Shannon in the general
theory of communication. The model proposed within that theory is shown below.
Figure 7
general schematic of communication [25]
We translate some of the sections figure 7 entire conceptual
language we can say that the information source and destination something
capable of storing information i.e. system. That the means of moving the
information from that source to the target is going to be a process or
something capable of realising a process such as a system. Finally, each of the
lines on this diagram represents what we would term a relationship and wherever
there is a relationship between information and process there is noise. When
noise seen as the possibility of information systems suffer from interference
or corruption. Interference is an issue with the transfer medium causing loss
of information and corruption as an issue with the copying mechanism causing a
loss of information.
Within a business context of this means is that once a model
is used to change the reality that each of the interfaces on it a be
relationships where information, process or systems interact there is the
possibility for the information travelling between each of those concepts to be
interfered with or corrupted. This means that anyone looking to ensure the best
possible model of the business should consider the controls required over
certain key interactions in order to mitigate the risks associated with
miscommunication.
Proposition six (Models must set Context & Boundaries) stated
that any modelling language would have to be placed within context boundaries and
the boundaries created should also be considered as areas for possible
miscommunication. That is, in the modern world, where we have process and data
architects acknowledged that those two professions are focussed on an
individual concept each. This means anyone looking to split apart the
fundamental concepts in a single modelling language or employ multiple modelling
languages which would split the overall model between professions should
consider the likely impacts that will have and ensure the necessary controls
are in place. In summary, silos within businesses whether they be technological,
or business solutions cannot be destroyed only reformed and their interfaces understood
and controlled.
Relationships are affected by noise
Not only does our view of the world change as accepted by
axiom one (Models must reflect and move with the Zeitgeist) but language also
determines how we see and think about the world. Given that in proposition six
(Models must set Context & Boundaries) it was deemed necessary that are
fundamental concepts have to be given context boundaries thusly follows that
any people using that language will have their worldview altered by it. This is
a view espoused by Ludwig Wittgenstein in his philosophical investigations
where he defines language games. Within the investigations, he pushes the view
that our language is our reality and each of our languages can be broken up to
a series of games. [32]
Where this becomes prescient as it points out as briefly
spoken about the end of proposition seven (Relationships are affected by noise)
is in creating interfaces whether they be between people or technology there
lies the possibility of miscommunication. Moreover, models generally used to
describe the past present and future which means that in those models are
considered to be a language broken up into games and that language constitutes
our reality then the models we choose will determine the solutions we go for.
This can be seen when looking through Microsoft as the guide
to architecture. What it specifies our particular styles of architecture or
technology that can be used when determining what products to buy from their
platform. When looking at each of the styles it quickly becomes a close
relationship between each of the models chosen in this particular investigation
and the architectural styles that will lead you to purchase particular
products. [26]
As noted in the original description of ArchiMate, it is a
modelling language focused around services and that each of the models it draws
over its layers and aspects of broken up into services that represent a
grouping of information and process akin to a definition of a system. Now one
of the styles Microsoft talks about is microservices which can be seen in the
diagram below.
Figure 8
microservice architecture from Microsoft Azure architecture guide [34]
What we can see is that there is a direct corollary between
the output of ArchiMate and a particular set of technologies readily available
from the cloud platform that uses the same type of language and modelling
structure. Thereby without further thought making it the most compelling technology
architecture to go for. [34]
This can be seen again when looking at BPMN the breaks the
world up into participants the activities they perform and then the messages
they share between one another. Looking below at the N-Tier Architectural style
in and then again season corollaries.
Figure 9
Microsoft Azure architecture guide N-Tier view [35]
Each of the layers performs a set of activities for the
lower level tasks which associate with one another either through a sequence of
events or using a message to relay information. Each component within this
architecture aligns again neatly with BPMN and may mean that people
inadvertently are driven towards technological choices that align with the
outputs of BPMN.
Our language determines our reality in the modelling
language we use to describe a business may result in its technology.
A Modelling Language may determine a Business’s technology
There was not time to give an in-depth investigation into
each individual concept and the different ways in which the concept could be
split up and categorised. It may be that within each concept there are also
crucial types of each which cannot be removed.
There are far more than just the three models used to
determine each of the fundamental concepts within this report. An extended
report or research looking at additional models and if they adhere to the
fundamental concepts devised will add any missing concepts further support the
existing concepts.
There is the possibility that process and systems could be
considered a relationship between states of information. This would limit the number
of fundamental concepts to two but might have the knock-on effect of vastly
complicating the creation of a particular context and boundary for a model that
could be used by a business.
The investigation into philosophical and scientific examples
for each of the concepts did not exhaust all the possible knowledge available
in a further investigation into more in-depth answers from each would either
support or potentially invalidate the definitions and propositions.
In conclusion the fundamental concepts of modelling that
must exist in an environment where we continuously update our knowledge that
only has finite resource to deal with infinite concepts which are information,
processes, systems and relationships. Where each of these can be grouped
together to describe something within a model. Where that model must be
governed by a set of rules that would make up a modelling language. Where those
rules are created to add context boundaries to that model. Interfaces within
that model are subject to noise and must be considered for control depending on
their importance. Those interfaces exist not only within the business & technical
solution but the practitioners who create the models. That any modelling language
will necessarily affect the likely outcome of technology used to realise the
models created with that language.
|
[1]
|
Open Architecture
Group, [Online]. Available:
http://www.opengroup.org/subjectareas/enterprise/archimate-overview.
[Accessed July 2018].
|
|
[2]
|
Object Management
Group, “BPMN,” [Online]. Available: http://www.bpmn.org/. [Accessed July 2018].
|
|
[3]
|
Object Management
Group, “UML,” [Online]. Available: http://www.uml.org/what-is-uml.htm.
[Accessed July 2018].
|
|
[4]
|
“Oxford
Dictionary,” [Online]. Available: https://en.oxforddictionaries.com/definition/data.
[Accessed July 2018].
|
|
[5]
|
“Oxford
Dictionary,” [Online]. Available:
https://en.oxforddictionaries.com/definition/information. [Accessed July
2018].
|
|
[6]
|
P. Adriaans, 2012. [Online]. Available:
https://plato.stanford.edu/entries/information/#BuiBloModTheInf. [Accessed July 2018].
|
|
[7]
|
P. Adriaans, 2012. [Online]. Available:
https://plato.stanford.edu/entries/information/#DevInf. [Accessed July 2018].
|
|
[8]
|
T. O. A. Group,
“Archi Specification,” 2012, pp. 12-13.
|
|
[9]
|
Open Management
Group, “Business Process Modeliing & Notation,” 2011, p. 60.
|
|
[10]
|
Open Management
Group, “Uniform Modeling Language,” 2017, p. 12.
|
|
[11]
|
O. Dictionary.
[Online]. Available: https://en.oxforddictionaries.com/definition/process.
[Accessed July 2018].
|
|
[12]
|
J. Seibt, 2017. [Online]. Available:
https://plato.stanford.edu/entries/process-philosophy/. [Accessed July 2018].
|
|
[13]
|
The Open
Architecture Group, “Archi Specification,” 2012, p. 14.
|
|
[14]
|
The Open
Architecture Group, “Archi Specification,” 2012, p. 17.
|
|
[15]
|
Open Management
Group, “Business Process Modelling & Notation,” pp. 53-54.
|
|
[16]
|
Open Management
Group, “Business Process Model & Notation,” 2011, p. 59.
|
|
[17]
|
Oxford Dictionary,
[Online]. Available: https://en.oxforddictionaries.com/definition/system.
[Accessed July 2008].
|
|
[18]
|
N. Weiner,
Cybernetics, 1961.
|
|
[19]
|
The Open Architecture
Group, “Archi Specification,” 2012, pp. 13-14.
|
|
[20]
|
Object Management
Group, “BPMN Quick Guide,” [Online]. Available:
https://www.bpmnquickguide.com/view-bpmn-quick-guide/. [Accessed July 2018].
|
|
[21]
|
Oxford Dictionary,
[Online]. Available: https://en.oxforddictionaries.com/definition/relationship.
[Accessed 2018 July].
|
|
[22]
|
M. a. S.-G. S.
Martinez, “Logic and Information,” Stanford Encylopedia of Philosophy, p.
1, 2018.
|
|
[23]
|
The Open
Architecture Group, “Archi Specification,” 2012, p. 22.
|
|
[24]
|
Object Management
Group, “BPMN SPecification,” p. 29.
|
|
[25]
|
“Oxford
Dictionary,” [Online]. Available:
https://en.oxforddictionaries.com/definition/model. [Accessed July 2018].
|
|
[26]
|
Open Architecture
Group, “Archi Specification,” 2012, p. 4.
|
|
[27]
|
J. S. Mill, On
Liberty, 1859.
|
|
[28]
|
D. S. Odenwald, “Stanford University,” [Online]. Available:
https://einstein.stanford.edu/content/relativity/a11332.html. [Accessed
July 2018].
|
|
[29]
|
J. J. Song,
“Linguistic Typology,” 2018, pp. 39-60.
|
|
[30]
|
J. J. Song,
“Linguistic Typology,” 2018, p. 141.
|
|
[31]
|
J. J. Song,
Linguistic Typology, 2018.
|
|
[32]
|
L. Wittgenstein,
Philisophical Investigations, 1953.
|
|
[33]
|
Microsoft, “Azure
Architecture Guide,” [Online]. Available:
https://docs.microsoft.com/en-us/azure/architecture/guide/architecture-styles/.
[Accessed July 2018].
|
|
[34]
|
Microsoft, “Azure
Architecture Guide Microservices,” [Online]. Available:
https://docs.microsoft.com/en-us/azure/architecture/guide/architecture-styles/microservices.
[Accessed July 2018].
|
|
[35]
|
Microsoft, “Azure
Architecture Guide N-Tier,” [Online]. Available:
https://docs.microsoft.com/en-us/azure/architecture/guide/architecture-styles/n-tier.
[Accessed July 2018].
|
|
[36]
|
B. Spinoza, Ethics,
1677.
|
|
[37]
|
C. E. Shannon, “A
Mathematical Theory of Communication,” The Bells Systems Technical
Journal, p. 2, 1948.
|
|
[38]
|
T. O. Group,
“Archimate Specification,” 2012, p. 29.
|
|
[39]
|
The Open Group,
“ArchiMate Specification,” 2012, p. 8.
|
|
[40]
|
Object Management
Group, “Uniform Modelling Language,” p. 14.
|
