Professional Documents
Culture Documents
a r t i c l e
i n f o
Keywords:
Informatics
Information
Information technology
Universals
Ancient Sumeria
a b s t r a c t
The term informatics is used as an umbrella term to stand for the overlapping disciplinary areas of information systems, information management and information technology. The current paper is part of a
series which documents an overarching attempt to develop a clearer and more sophisticated systematics
for the area. It examines one of the foundation concepts of informatics that of information and aims to
provide a denition for this concept based upon ideas from semiotics and communication theory. For this
purpose we introduce the concept of a sign-system and consider the role such a system plays in human
communication. We also highlight the fundamental difference between a communication system and an
information system. To help ground our discussion and provide a necessary distance from the presentday concern with digital computing and communication networks we engage with the historiography of
information. We consider the use of information in Neolithic times and describe the case of clay tokens in
Ancient Sumeria as one of the earliest examples of information representation and manipulation. Examination of this case allows us to propose a number of universal features of information and information
technology.
2008 Elsevier Ltd. All rights reserved.
and that the use of such signs involves human interpretation. The
examples also demonstrate that information is inherently bound
closely with numerous phenomena language, action, logic and
technology; to name but a few. Hence, an understanding of the place
of information in human organization demands a multi-layered or
multi-levelled perspective.
3. Signs and communication
To help us construct a fruitful denition of the concept of
information we build upon Stampers (1973) seminal work on Information in Business and Administrative Systems published close to 30
years ago. In particular, we use the framework of semiotics to help
understand the multi-faceted nature of information. The study of
information is necessarily a boundary-spanning activity since information inltrates a multitude of areas; all of which are important
to organizational informatics.
Broadly, semiotics or semiology is the study of signs. Signs are
seen as the core element of concern serving to link issues of human
intentions, meaning, the structure of language, forms of communication transmission, data storage and collaborative action.
We would argue that the concept of information revolves around
this concept of a sign; being anything that is signicant. In a sense,
everything that humans have or do is signicant to some degree.
Sometimes not having or doing anything is regarded as signicant. The world within which humans nd themselves is therefore
resonant with systems of signs.
Steven Pinker (Pinker, 2001) argues that our genetic makeup
predisposes humans to be excellent manipulators of sign-systems.
A sign-system is any organized collection of signs. Everyday spoken language is probably the most readily accepted and complex
example of a sign-system. Signs however exist in most other forms
of human activity since signs are critical to the process of human
communication and understanding (Stamper, 1973).
Mention of linguistics and language should not however restrict
our conception of a sign to spoken language. For example, and as
indicated in the examples above, in traditional forms of face-toface communication, humans communicate through non-verbal as
well as verbal sign-systems; colloquially referred to as body language (Morris, 1979). Hence, humans impart a great deal in the
way of information by facial movements and other forms of bodily gesture. Such gestures are also signs. An example here is the
hand gesture in which the rst two ngers are used to create a V.
If this gesture is made with the palm facing toward the person(s)
communicated with, it might typically signify victory or possibly
peace. The gesture made with the knuckles facing towards the person(s) normally means something entirely different, at least in the
UK!
Note the link between the words sign and signicant in English;
they have the same root. Hence, the signicance of signs cannot be
divorced from people. Different people nd different things signicant. Many such differences in interpretation are due to differences
in the context and culture of communication.
Charles Morris (Morris, 1964) originally proposed three
branches of semiotics pragmatics, semantics and syntactics. In
(Stamper, 1973) signs and sign-systems are considered in terms of
four inter-dependent levels, layers or branches of semiotics: pragmatics, semantics, syntactics and empirics. These four layers serve
to connect the social world on the one hand with the technical
world on the other (see Fig. 1) (Stamper, 2001). Hence, this layered
model, which is sometimes referred to as the semiotic ladder, serves
to represent the concept of information as necessarily a sociotechnical phenomenon interposing between three different levels
of system of interest to organizational informatics: activity systems,
information systems and ICT systems. It also provides greater clar-
or commodity. Hence, a given token signied both a type of commodity and the quantity of this commodity.
Schmandt-Bessarat (1996) suggests that the way in which
tokens signied goods and quantities changed over the millennia
as a reection of social change. At rst tokens were used to signify
staple agricultural commodities. Later, tokens were used to signify
goods manufactured by craftsmen in cities. In other words, simple tokens were initially used to signify unprocessed commodities.
Over time, complex tokens evolved to signify processed commodities.
When the token system came about in the pre-Sumerian era,
circa 8000 bc, the rst tokens consisted mainly of abstract shapes
formed in clay such as cones, spheres, tetrahedrons, discs and cylinders. She suggests that these earliest tokens probably recorded the
most basic staples such as quantities of grain and livestock. These
so-called plain tokens continued to be used until the end of the
third millennium bc (Fig. 3).
In about 4400 bc, what Schmandt-Besserat refers to as complex
tokens started appearing in the early cities of Sumer. These tokens
consisted of new shapes and the use of incised markings (Fig. 4).
She proposes that these complex tokens stood for nished products,
such as bread, oil, perfume, wool and rope, and for items produced
in workshops such as metal, bracelets, types of cloths, garments,
mats, pieces of furniture, tools and a variety of stone and pottery
vessels. It is noteworthy that these complex tokens did not replace
the earlier plain tokens but were merely added to form a larger
system of signs.
Schmandt-Bessarat (1996) categorises tokens into 16 main
types based primarily upon shape (Fig. 5). These include cones,
spheres, discs, cylinders, tetrahedrons, ovoids, rectangles, triangles, biconoids, paraboloids, bent coils and ovals/rhomboids. She
also identies a number of sub-types based upon variations in
sizes and markings. For example, cones, spheres, disks and tetrahedrons are typically represented in two sizes small and large.
Many shapes also have incised markings consisting of incised lines,
notches, punches and pinched appendices. The numbers assigned
to each type in Fig. 5 represent the number of sub-types in each
category identied by Schmandt-Bessarat.
The evolution of the token system therefore appears to reect an
ever-increasing need for accuracy. For example, in tokens dealing
with livestock, early plain cylinders and lentoid disks apparently
stood for heads of livestock. In contrast, in the fourth millennium
bc complex tokens signied different species of livestock such as
fat-tail sheep, the sex of the animal such as ewe and the age of the
sheep such as lamb. This increase in the number of token types and
sub-types, which occurred in large cities of Sumer about 3500 bc,
seems therefore to indicate a concern for more precise data and
consequently a more sophisticated form of record-keeping.
The clustering of tokens in groups varying in size from two to
about one hundred tokens at various archaeological sites suggests
that a given token typically signied a small quantity of a given commodity; probably signifying a one-to-one correspondence between
a given token and one unit of the commodity signied. Hence,
one jar of oil could be represented by one ovoid, six jars of oil
by six ovoids, and so on. There were apparently only a few tokens
that stood for a collection of items. For example, the lentoid disk
probably signied a ock of perhaps ten animals. This meant that
the token sign-system did not allow the user to express numbers
abstractly. In other words, there was no token for the concepts of
one, two and three (the concept of cardinality) independently of the
commodity counted.
In the early fourth millennium bc two mechanisms were also
devised for aggregating a group of tokens. Tokens were either tied
together with a piece of string or were enclosed in hollow containers of clay. We shall focus on the latter form of aggregation, which
Schmandt-Besserat refers to as clay envelopes (Fig. 6). Initially, it
appears that these clay envelopes were used as a means of imprinting the seals or signatures of the parties engaging in the economic
transaction being recorded. Such signatures were hence probably
used to signify ownership, obligation or authority. After a period of
time, the impressions of the tokens contained in the envelope were
also impressed in its outer layer. After the envelope was completed
it was baked, making the record persistent or difcult to alter.
6. Accounting in Ancient Sumer
Schmandt-Bessarat (1992) hence argues that tokens were a conceptual leap for human-kind, constituting a new means of encoding
data. The essential purpose of tokens was to keep records of things:
an accounting. The shape and markings of the token signied the
thing about which the record was made. Hence, a conical type
of token probably stood for a small measure of grain, while a
spherical token stood for a larger measure of grain. In contrast, a
cylindrical token probably stood for a domesticated animal, a tetrahedron token for a unit of labour and an ovoid token for a jar of
oil.
Some have argued that clay tokens are not actually the earliest evidence of human symbolism. Marshack (2003), for instance,
suggests that symbolism such as scratches on antler horn from the
Upper Palaeolithic (35,00010,000 bc) period may constitute some
of the earliest use of a tally, related perhaps to calendar events such
as the gestation period of a horse or the phases of the moon.
However, clay tokens appear to be a distinct innovation in a
number of ways over the use of tallies.
First, they were entirely man-made artefacts compared to earlier symbolism which involved modifying natural objects. This
meant that distinct token types could be associated with a discrete
and unequivocal meaning that was easily communicated between
human actors. A tally in contrast is meaningless without communication of the context in which it is created. This enabled tokens
to provide a far more exible form of signication on a number of
counts.
Second, the forms of the token-system were relatively easy to
create and duplicate. Clay was a commonly available material which
required no special skills or tools to work. In most cases creation
of a token involved merely rolling a small lump of clay between
the palms of the hands and pinching it between the ngertips in
various ways.
Third, the tokens as a whole constituted a sign-system. This
allowed the classication of things. The token-system was also
open-ended in that the signication of new types of commodities
merely required the addition of a new type of token to the tokensystem (a new shape or an existing shape with new or distinctive
markings) (Rudgley, 1999).
Fourth, since tokens of different kinds and in different quantities
could be stored together, it has been proposed that they were used
as one of the earliest forms of accounting record (Mattesich, 2000).
Such records precluded the need for any individual to memorise
accounts. They could be referred to and understood at any future
time by someone who knew what the token represented.
Fifth, it has been proposed that tokens were independent of phonetics. Hence, the Sumerian token system could be regarded as one
of the few examples of semasiographic writing. Although most
existing systems for writing natural languages are based on a correspondence between the written sign and some element of the
spoken word, a system of signs does not have to replicate speech to
communicate narrative. Therefore, people speaking different languages or dialects found it easy to adopt this system based on
physical artefacts. The evidence of extremely wide distribution of
such tokens throughout the Near East in Neolithic times seems to
support this inference.
Sixth, because of the one-to-one correspondence between the
token and the commodity represented, it was relatively easy to
apply operations such as addition or subtraction within the system.
However, when the numbers of things to be counted was large, such
operations could be time-consuming and tedious.
Mattesich (1989) describes the clay envelope with its enclosed
tokens as equivalent to modern notions of a personal account
representing that portion of total assets (or equity) invested in
a particular debtor. . .Consequently the sum-total of the various
tokens in a particular envelope. . .stood for that part of the creditors wealth lent to a debtor. The token system therefore contains
elements for representing both the physical output of goods from
one place and input into another place. It also appears to have been
used to represent social relations of ownership and debt.
Tokens were apparently not re-used once made. They were disposed of once the transaction they represented was concluded.
Evidence from archaeological investigation suggests that the use
of tokens for keeping accounts followed an annual cycle of activity
associated with agriculture. Tokens appear to have been discarded
after the harvest and threshing, when the crops would be stored.
This suggests that economic transactions were made in the course
of the year to be completed at the time of the harvest. If this was
so, the usual length of keeping accounts in archives was less than a
year.
7. Association with religion, social hierarchy and the
development of writing
Rowland (2003) summarises a number of suppositions from
the historiography of communication, particularly as embodied in
the work of Innis (1950). The rst supposition is that all communication technologies (or what we would refer to as information
and communication technologies) are extensions of basic, innate
human communication capabilities. The second supposition is that
different communication technologies impact upon the cognitive
structures of human beings over time, inuencing changes in social
organization. The work of Schmandt-Bessarat (1996) supports such
reasoning. In her research she proposes an inherent association
between the use of tokens for record-keeping purposes and the
rise of social hierarchy, centralised religious places of worship and
the development of writing in Ancient Sumeria. Tokens and clay
tablets functioned as an extension of the human brain to collect,
manipulate, store and retrieve data. In turn, processing an increasing volume of data with more complex tokens brought people to
think with greater abstraction (Schmandt-Bessarat, 1992).
10
11
Human communication is therefore a necessary but not a sufcient condition for an information system. The essence of an
information system is embodied in the concept of the record and
the activity of record-keeping. The record acts as an external and
persistent memory of something. The record signies something
intended by the maker of the record that can be interpreted by the
reader of the record at some other time and possible place. Using
such artefacts, the dialogue or discourse of human agents can occur
across time and possibly space.
We have used the discipline of semiotics and its concept of a
sign as a means of providing greater clarity to the concept of information. Semiotics is devoted to the study of signs and consists of
four sub-areas. The levels of pragmatics, semantics, syntactics and
empirics represent various facets of the concept of a sign that span
between the social and the technical.
The focus of pragmatics is on the intentions of human agents
underlying communicative behaviour. In other words, intentions
link language to action. The case described in this paper indicates
that record-keeping as a formal language appears to pre-date the
invention of written language. There is evidence of this in other cultures such as the Inca where a form of record-keeping based upon
knotted strings emerged in a society without a written language
(Urton & Brezine, 2005).
Human memory is sufcient to support cooperative and simple
activities between individuals in small communities. As communities grow, activities, particularly those reliant on economic
exchange, need to take place between strangers and generally
are more complex in nature, reliant typically on some division
of labour. Records compensate for the limitations of individual
human memory and extend it into social memory. Records of
economic transactions institutionalise memory of past economic
exchanges and the obligations placed upon individuals engaged in
such exchange. It is also proposed that accurate record-keeping
is critical in establishing and sustaining trust between strangers
engaging in economic exchange. Records account not only for the
types and quantities of commodities exchanged, they are also
important for supporting social relationships such as ownership
and debt.
Semantics is concerned with the meaning of a message conveyed in a communicative act. Semantics considers the content of
communication and is the study of the meaning of signs. Semantics can be considered as the study of the link between symbols
and their referents or concepts. In contrast, syntactics is concerned
with the formalism used to represent a message. Syntactics studies
the form of communication in terms of the logic and grammar of
sign-systems. In other words, syntactics is devoted to the study of
the form rather than the content of signs.
John Dewey (Dewey, 1916) suggested that a word is three things:
a fence, a label and a vehicle. The same could be said more generally
of the concept of sign. A sign is a fence in the sense that it sets a
conceptual boundary around some thing and is used to distinguish
one thing from another. A sign is a label in that it acts as a convenient reference standing for something else. Finally, a sign is a
vehicle in the sense that used with other signs as a language it
is a means for describing and debating with the world as well as
acting upon it. Pragmatics is particularly interested in the sign as
a vehicle as a means for supporting action. Syntactics is particularly interested in the sign as a fence how conceptual boundaries
are created with symbols. Semantics is particularly interested in the
idea of the sign as label and as such is the study of what symbols
refer to.
A simple model which encapsulates some of the essence of
semantics and syntactics is one in which a sign is broken down
into three component parts which are frequently referred to collectively as the meaning triangle (Fig. 7) (Ogden & Richards, 1923).
12
The designation of a sign refers to the symbol (or collection of symbols) by which some concept is known. The extension of a sign refers
to the range of phenomena that the concept in some way covers.
The intension of a sign is the collection of properties that in some
ways characterise the phenomena in the extension, and is the idea
or concept of signicance.
In this conception, symbols are equivalent to data. A datum, a
single item of data, is a symbol or a set of symbols used to represent something. Information particularly occurs in the stands-for
or intentional (Searle, 1970) relations between the symbol (designation) and its concept (intension) and the symbol and its referent
(extension).
The Sumerian clay token is a classic example of a sign. The token
itself represents the designation of the sign the symbol. A token
signies two concepts through its existence, shape and markings
(intension): the quantity of some commodity and the type of commodity. Hence, there was meant to be a one-to-one correspondence
between the designation and extension within the sign-system.
For example, one token of a given type represented one item of
a commodity such as a measure of grain or oil.
The collection of both plain and complex tokens constituted as
a whole this sign-system. The physical makeup of the token themselves and the ways in which the features of each token allowed the
user to distinguish between one type of token and another constituted its syntax. Given the demise of the human community that
used this particular sign-system, the semantics of tokens is subject
to much interpretation. However, Schmandt-Besserat believes that
there is sufcient evidence for a number of associations such as
those illustrated in Fig. 8 for particular types of plain and complex
token.
Empirics is the study of the signals used to carry a message; the
physical characteristics of the medium of communication. Empirics traditionally considers problems solely of data communication,
particularly the study of communication channels and their characteristics. However, empirics should also logically concern the issue
of data representation and storage based around the concept of a
record.
The term universe of discourse (UoD), domain of discourse or
ontology (Kishore, Sharman, & Ramesh, 2004) is sometimes used
to describe the context within which a group of signs is used continually by a social group or groups. For work in informatics, at the
level of empirics, it is important to develop a detailed understand-
elements and les are the data structures within this data model.
Although this data model assumed some signicance with the rise
of the modern ofce and bureaucracy (Weber, 1946), we would
argue that the le-based, sometimes referred to as the recordsbased, data model has existed for many thousands of years in
numerous distinct human civilizations. In a sense, records by their
very nature involve the use of signs to act as a persistent signication of something. By persistence we mean that a communication
is encoded in some reied form which allows it to be transported
through time and space. Records are typically collected together in
the data structure of a le. Collecting records together in a particular
le normally implies some association (semantic intent) between
these data elements.
One can clearly argue that accountants in Ancient Sumeria used
an inherently le-based data model based around the creation and
manipulation of physical tokens. Clay tokens of various types constituted the data items in this data model. The analysis conducted
by Schmandt-Bessarat indicates that the variety (Beer, 1972) in this
sign-system is 492 (the sum of the number of sub-types in Fig. 5).
Hence, there are roughly 500 different things that could be represented as data items within this data model.
Sometimes, a collection of such data items were collected
together and enclosed in a clay envelope. This envelope was xed
in baking as a data element or record of an economic transaction. It could also be argued that the co-location of tokens and clay
envelopes within some location such as a temple precinct equated
to the creation of a data structure or le which served perhaps to
record information such as the surplus of a particular harvest stored
for later use.
Schmandt-Bessarat cogently summarises her position in the
following quote: The conceptual leap that revolutionised communication was the creation of a set of symbols with specic shapes
and endowing each shape with a discrete meaning. The tokens were
modelled in striking, geometric shapes that were easy to recognise. The forms were simple and easy to duplicate. The counters
were systematically repeated, always carrying the same meaning.
The dozens of token shapes constituted a code of dozens of interrelated concept symbols concerning goods and commodities. Later,
with the development of a repertory of markings, the code grew to
hundreds of concept symbols. The system made it possible to deal
concurrently with multiple kinds of data, thus allowing the processing and communication of a volume and complexity of information
never reached previously (Schmandt-Bessarat, 1992).
If we replace Schmandt-Bessarats use of the term concept symbol with sign in this quote we would argue that the four levels of
sign described above, relate directly to three levels of system important to informatics. Human activity systems are social systems
and hence typically interact with the semiotic level of pragmatics.
Signs are used within such activity systems to support collaborative and coordinated action. In one direction they are used to encode
intentions. In the other direction they are used as key inputs into
decisions made about action. An information system is a communication system used to support a given activity system and as such
is mainly located against the semantic and syntactic levels of signs.
The output of an information system is therefore information; information systems use agreed systems of signs to represent meaning.
An ICT system is a designed system of artefacts used to collect, store,
process and disseminate data. Since the output of an ICT system is
data, such systems are located mainly at the technical and empirics
level of signs. ICT is concerned with the physical representation of
signs for storage, transmission and manipulation.
Within the Sumer case, the activity system of concern involved
the storage and distribution of wealth produced by a settled
community. Tokens were used within an information system to
represent the accumulation of agricultural and manufactured prod-
13
14
Burgoon, M., Hunsaker, F. G., & Dawson, E. J. (1994). Human communication. Thousand
Oaks, CA: Sage.
Castells, M. (1996). The rise of the network society. Mass.: Blackwell.
Checkland, P. (1999). Soft systems methodology: A thirty year retrospective. Chichester:
John Wiley.
Clark, H. (1996). Using language. New York: Cambridge University Press.
Crowley, D., & Heyer, P. (Eds.). (2003). Communication in history: Technology, culture,
society. Boston: Pearson Education.
Currie, W. (2000). The global information society. Chichester: John Wiley.
Dewey, J. (1916). Essays in experimental logic. Chicago: University of Chicago
Press.
Garnkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Gregor, S. (2006). The nature of theory in information systems. Management Information Systems Quarterly, 30(3), 611642.
Harris, R. (1986). The origin of writing. London: Duckworth.
Hobart, M. E., & Schiffman, Z. S. (1998). Information ages: Literacy, numeracy and the
computer revolution. London: John Hopkins University Press.
Innis, H. (1950). Empire and communication. Toronto: University of Toronto Press.
Kishore, J., Sharman, R., & Ramesh, R. (2004). Computational ontologies and information systems. 1. Foundations. Communications of the Association for Information
Systems, 14, 158183.
Klein, R. G. (2002). The dawn of human culture. New York: John Wiley.
Kling, R., & Allen, J. P. (1996). Can computer science solve organisational problems?
The case for organisational informatics. In R. Kling (Ed.), Computerisation and
controversy: Value conicts and social choices. San Diego: Academic Press.
Marshack, A. (2003). The art and symbols of ice age man. Communication in history:
Technology, culture and society. Boston: Pearsn Education.
Mason, R. O., Mckenney, J. L., & Copeland, D. G. (1997). An historical method for MIS
research: Steps and assumptions. MIS Quarterly, 21(3), 307319.
Mattesich, R. (1989). Accounting and the inputoutput principle in the prehistoric
and ancient world. ABACUS, 25(2), 7484.
Mattesich, R. (2000). The beginnings of accounting and accounting thought. Routledge.
Morris, C. (1964). Signication and signicance. Cambridge, MA: MIT Press.
Morris, D. (1979). Manwatching: A eld guide to human behaviour. London: Harry N.
Abrahams.
Ogden, C. K., & Richards, I. A. (1923). The meaning of meaning. London: Routledge and
Kegan Paul.
Pinker, S. (2001). The language gene. Harmondsworth, Middx: Penguin.
Porter, M. E., & Millar, V. E. (1985). How information gives you competitive advantage.
Harvard Business Review, 63(4), 149160.