Semiotics and the Paradoxes of Agency

0. Introduction
The internet is praised and blamed for many things. It might be useful to explore some of the fundamental paradoxes of agency (and presence) that semiotics affords, so that we might be able to understand things a bit better. Here are some thoughts ...

Semiotics: the capability to forge relationships between experience and communicative action, and let it loose on the world about you – is the basis for relationships, inter-subjectivity, self-consciousness, innovation and creativity, presence and power.

We are uniquely creative and destructive animals, because of our ability to create new ways of perceiving, interpreting and acting on the world, capture and accumulate all this, and pass it on from generation to generation. In the process we create our own social reproduction, which may overshadow our given, biological reproduction (in more than one sense). We can, in ecological terms, change and adapt our own evolution (or ‘devolution’ as the case may be).

This is a dilemma, which is most obvious in our ‘edge of chaos’ behaviour as a species, still living in the cold shadow of our global nuclear balance of terror, or MAD (Mutually Assured Destruction), and on the edge of disastrous, unstoppable, climate change.

In order to understand how we have come to where we are, it is useful to explore the many layers of paradoxes that underlie our capability to perceive, interpret and act on the word.

This is an overview of the socio-genesis of some of the key elements of our current situation, with some thoughts on the paradoxes involved at each stage (in indented text).

1. Signs
Semiotics, the meaning of signs (from hand-axes to music, mathematics, and complicated theories) is based on use, and every use becomes a sign of itself (Roland Barthes 1977). The more a sign is used, and the people use it, the more it becomes a recognised sign amongst a particular community, and gains meaning and currency.

2. Ecologies of Meaning
Signs, in the semiotic sense, are not physical objects – like ‘road-signs’, or even printed words. They are relationships between an experience and a communicative &/or interpretative action - you use them to capture, convey, point to, explain, question, or describe these relationships. Signs are ecological: they are used to distinguish difference – difference in experience, in interaction, and from other signs. Signs, like genes, function in an ecology of systems of difference – differences which change, adapt and emerge as the ecology unfolds (or collapses).

3. Flexibility
All signs are to a greater or lesser extent flexible. They can be changed, or used to replace other signs that don’t work as well. As Umberto Eco succinctly put it, a signs is something that can used to lie. That means that just because someone has a name for something does not prove that it exists – it might or might not, it might just be a fantasy, you will only find that out by using it, with others.

...#1: But that doesn’t stop some people from believing - and even enforcing the belief - that certain physical, virtual or metaphysical things exist. Which means that all relationships on the internet might be with avatars, or with people deliberately deceiving you that they are someone else, so all signs should initially be taken with a pinch of salt.

4. The paradox of meaning
Linguistic signs are very flexible, as they are made up of a combination of arbitrary sound elements to form useful (i.e. meaningful) signs. Particular combinations of sound or graphic elements may have some correlation in sensory experience, they may not. But they are all in principle arbitrary: ‘god’ and ‘dog’ could be interchanged, if a community of users agreed on that. So all signs are in principle both arbitrary and conventional, they may be 'motivated' by sensory experience, but they are at the same time open: they provide new opportunities for users, new combinations, and new relationships to experience, to things and to people (i.e. new affordances[i], in ecological psychology).

...#2: This means that you can create and use signs anyway you want to. However, you have to convince others that it makes sense for them to use them in the same way too. So your agency, viz: your capability for independent initiative, is on the one hand infinite, but on the other hand it limited is in practice by the agreement you can reach within a group of other people who share that use (and meaning) with you.

...#3: This in turn means that all semiotics can be used for inclusion and exclusion, for sharing or for division - from the slang in the playground to jargon in economic policy. Just more layers of systems of difference.

5. Indicative to Predicative
Individual signs can serve many purposes, but they are somewhat limited to description, and to ‘pointing’ – to indicative functions. It is when they are combined with other signs to form instructions, sentences, and syllogisms, which link functions between subjects and objects (via predicates or verbs in sentences) that they can start to potentially take on more substantive meaning, and a ‘life’ of their own, as predicative statements. The instructions between a shepherd and a sheep dog, for instance, are a fascinating borderline category between indicative and predicative communication, and between animal and human semiotics.

As this progresses to more complicated combinations of signs - such as syllogisms, aesthetic works, equations, program code, etc, they are able to function in a different way, at a different level: i.e., as micro-assemblages of stand-alone signs that can be exchanged and used by quite different people.

...#4: This means that as soon as you create these packages of signs, they can be used independently of you – anyone else can in principle use them in different contexts, and remix them for different purposes too. This in turn means that as your agency (you capability to create and use more complex semiotic assemblages) becomes more powerful, so too does the potential for those assemblages to take on a life of their own increase too (outside your control) which is the most basic Pandoran[ii] paradox: you cant put the signs and texts ‘back into the genie’s lamp’ (to mix metaphors) from signs to texts to the genetic code for Anthrax (for instance).

6. Predictive knowledge
We can create and use signs for whatever we want (as longs as we can find, persuade, or force enough people to agree with us).

We can use signs and assemblages of signs to describe facts, and to create fact-based knowledge. We do this is by developing signs that describe things in a completely disinterested way. To do that, we have to strip out all context, and any particular subjects from our descriptions of experience, and/or the patterns we are describing. That gives us objective, predictive knowledge. However, this is always provisional, until we find a better sign or assemblage of signs. Predictability is only as good as it’s last use.

...#5: We can always cheat, of course: we can force people to accept the way we ‘predict’ - and want - things to be, but that uses coercion to displace logic, and compulsion to displace cooperation, and it’s only ‘predictable’ as long as a person, agency, Church or State is around to enforce it.

7. Commodified signs
Commodified signs (which are often expressed and/or valued in numbers or mathematics) are signs or assemblages of signs that are used primarily for exchange – for use by anyone, anywhere, anytime – they have ‘currency’, literally – they can be moved and passed around, ‘freely’ (as in the ‘frictionless’ financial economy).

Commodified signs are also paradoxical, as they depend for their initial exchange value on their immediate or potential use, even though they can take on a life of their own, and be traded, for instance, in currency markets, or even futures currency markets.

They are abstract, and derivative, as they are stripped of all provenance – i.e. context, subjects (or users) and subjectivity. Money is the prime example, and Marx’s analysis - of the paradoxical relationship between exchange value and use value - can be applied quite broadly to other types of commodified signs.

There are three other key types of commodified signs: science, bureaucracy, and representative democracy, all of which deal in commodified signs. Science can be used by anyone, which is why superpowers go to such length to prevent some people (e.g. Iran, but not Israel) from developing fissile materials for nuclear arms); the rules of bureaucracy (and the laws that back them up) are designed to be impartial – the symbol of justice (in the West) is a blind woman holding two scales) for that reason; and representative democracy gives politicians like George W. Bush the political capital to "spend as I see fit" (as he proclaimed after winning his second term as President). Once commodified signs are established, they may become more and more derivative (e.g. the infamous multi-derivatives of the speculative financial market collapse in 2008), and they are can be endlessly accumulated (in Billionaire’s personal bank accounts, or in sovereign investment funds).

...#6: Commodified signs (and commodified semiotics – the systematic use of commodified signs) take the form of ‘disinterested’ signs. However, the use of disinterested signs, which appear to be ‘free’ in many senses of the word, actually depends on who you are, and how you are sited, sighted and cited: in short, what communities you belong to, and your status within them. It is not true, for instance, that anyone can become a member of Parliament, a nuclear (arms) scientist, or a currency trader. Over and above certain basic technical capability, you may require particular social and political capital, not to mention a security clearance and even a few international treaty agreements, none of which are evenly distributed across the globe, or freely accessible.
This paradox, #6, (and #7/8, below), are thus extrapolations of paradox #4.

...#7: Commodified signs (expressed as they are in, numerical, mathematical, or computer code forms) are derivative – they are abstracted from everyday semiotics. As Marx (as semiotician) says, money transforms relationships between people into relationships between things – and, nowadays, virtual things, as the pretence that money is a physical object has now largely been discarded. It might be useful to call them meta-semiotics (Williams 2005), as they function apart from, and even ‘outside of’ everyday semiotics.

...#8 …and this ‘meta’ abstraction of commodified semiotics makes them available to a smaller and smaller number of more and more powerful people, as the more abstract they are, the more complicated it is to realise your (potential) agency to use them – and the easier it is for the community of people who can use them effectively, to use them for exclusion of others.

8. Emergent Semiotics
Signs are relational, between (at least two, major) systems of differences, i.e. experience and communicative action. As Derrida (and many others, from De Saussure onwards) have pointed out, there is no point at which these systems of differences stop – there is no ultimate signifier – its differences all the way round.

So, at the level of everyday speech, signs resonate and/or dissonate with each other in a continuous slippage of meaning and use (what the French semioticians call parole), i.e. in an open ecology, and an open interplay of meaning.

#9: As signs become combined into assemblages of signs – sentences, syllogisms, equations, programs, media texts, etc., the paradoxes of structure kick in. It is then possible (not inevitable) that exclusion increasingly dominates over inclusion; orthodoxy dominates over sharing and creativity, and agency and freedom, and the conventional part of 'signs are arbitrary and conventional' can take on a decidedly political and partisan nature. In short, openness can end up feeding closure. This can in turn lead to the appearance of certainty - in science just as in religion.

...#10: However, hypertext (the internet protocols which provide the functionality for creating links between any sign on any website and any other signs, across the globe, almost instantly) has the potential to pierce the constraints of exclusion and orthodoxy, and to provide affordances for sharing, and for renewed agency and presence in the hyper-connected, playful ecology of interaction and meaning.

This hyper-semiotics (or emergent[iii] semiotics) provides a further ‘layer’ of semiotics, which functions alongside – and simultaneously interpenetrates - commodified (or meta-) semiotics, as well as everyday semiotics (of parole – in the broadest sense).

In doing this, it reinstates a kind of virtual parole – and the affordances of open, flexible, unpredictable, more creative communication, in a complex, adaptive, social ecology in which communicative and interactive agency is re-initiated and re-established at what Knorr-Cetina calls the micro-social level – an inversion of Foucault’s capillaries of power (in which power and agency is largely determined at the macro-social level, which then penetrates downwards to the ‘capillaries’ of the micro level).

(The generic term ‘social media’ is a reasonable approximation of what is meant by hyper-semiotics, and is probably more user-friendly and accessible to the general reader).

...#11: So, we can say that ‘social media’ provide radically new agency, particularly as they originate from the micro level – of a single Tweet, a single blog site, etc, - which can have impact on global scale. And because we are fast becoming a mobile-communication-device species, the incremental cost of this agency is effectively zero, it’s fast becoming just a by product of our wirelessly connected lives. That’s the good news.

...The paradox is that although it is true that social media is open and accessible, and that because most social media traffic is (en passant) - recorded conversation - users can remix, reversion, and repurpose each bit of it, creating exponentially more agency, there is also another side to it.

Namely, it is also true that all recorded communication traffic can in principle be hacked and vacuumed up by State surveillance agencies, and a range of other legitimate and illegitimate agencies, including terrorists, all of whom operate, for the most part, outside of democratic oversight, let alone control, and all of which are in the business, in some ways, of exercising control and limiting agency. It is also true that these same ‘grey agencies’ can increasingly take over your internet device, and turn it into an obedient node (or botnet) in their network, thus incorporating or pirating your agency to serve theirs.

As a species, we have to (continually) learn to adapt to our new freedoms.
..#12 And now war and assassination (including in countries like the UK that have banned capital punishment) is also a semiotic paradox, in the silences of war.
Comment, from Matthias Melcher: (28.09.16)
With responses from RW (aka dustcube) in pink ...

I found your Agency page fascinating and I learned a lot from it,
in particular how the assemblage of new language uses is linked
to an essential human capability, and how knowledge can be seen
as a 'disinterested' description of experience, and as
essentially predictive.

I like very much how you, as you go along, explain much of the
genesis of language. I understand that this is not your main
focus. But I think it would be a pity not to follow this path
right through to the little missing bits -- in particular in
your context of (resonant) knowledge.

Indeed, the trick is to do this without too much repetition.

So I would like to enumerate the points where your account
is very close to a full general explanation of language genesis. Thanks for all your input.

1. You begin with the Sign as a starting point that is already
there. However, many passages elsewhere talk so plausibly in
so much detail about the abstraction-like processes that may
have created these signs, and about the link between the sign
and the real world experiences. I think, the experiences can
then easily be interpreted as McGilchrist's right hemisphere,
and the drive to create a sign can be interpreted by the
fixing/ isolating/ grasping/ referencing/ intending/
representing/ wrapping/ collapsing aspects of McGilchrist's
left hemisphere.

Precisely. And ‘collapsing’ in particular is very useful, and it adds to the ‘meaning cloud’
of ‘stripped’, ‘abstracted’, and ‘revealed’ (in the sculpture example).
In this first stage of your section 1, this
would be not more than the statistical association of the
encounter with a certain thing in the world (say, a sweet red
berry) and the utterance of some sound by the human
(say "uh! uh!"). Assemblages follow later.

Yes, patterns of use are followed / accompanied by patterns of assemblages,
and cross-modal recognition (and then re-use) of the ‘fit between the patterns
of use and of an open collection of relevant signs (of all types).

2. Your characterization of this basic action as 'indicative'
reveals a lot of relevance for the early humans who became more
successful just by such indicative sharing/ encouraging to
imitate a successful peer and help them to recognize the same
useful experience. To illustrate this fundamental act, the
term 'pointing' seems a bit narrow. Maybe this is due to
my limited EFL command of your language, but it reminds me only
of the pointer finger of a parent, or the pointer stick of a
teacher, who directs the (focused, left-brained) attention of
the child to an object. Cooperative recognition and imitation,
however, must have been a much richer adventure for the early

Sure, ‘indicative’ is a much broader term in semiotics, but ‘pointing’
is a useful place to start to show (some of) what it includes.

Your frequent allusion of power and coercion reminded me of
the following:

In German, the term 'zeigen' (akin to 'Zeichen' = 'sign') is much
more powerful, and includes ideas of demonstrating (as in Downes's
saying "to teach is to model and demonstrate") and showing off
something, maybe a novel artefact. In the old, Germanic,
stages of the English language, the rich sense was probably
still present ("Proto-Germanic *taikijan "to show"", see ) but then it
became associated with coercion (zeihen = 'accuse') and was no
longer useful for agency.

Whew! So much to explore - ‘sign’ is useful as it covers so many things, but it is
also broken up into signifier (the mechanism or ‘agent’ of meaning) and signifier
(the thing, process, agent, institution, imaginary or otherwise) that it links to.
‘Signs’ are defined not as ‘things’ in semiotics, but as a dynamic relationship
between signifier and signified. (In American, or Peircian semiotics, this is done
a little differently - I prefer the above definition, which is de Saussurian, or European)

(BTW, maybe a similar development caused the word 'teach' to
sound much more authoritarian (and like a raised pointer finger)
than its German equivalent 'lehren' which literally means
'make so. learn, make so. track'. And the previous word for
danger, fær, was akin to experience and pirates and empirical,
while 'danger' is akin to dominance, i.e. danger arising from

So, the referent of the first signs might not have been
pointable objects yet -- while indeed the sign became just such
a pointed, shrunk 'handle', as McGilchrist explains.

3. Then when you talk about the composition of signs, your
example of predicates seems just one step too far already.
Between single part utterances and full sentences, lie a lot
of composite terms, whose creation may have been a great
example of what you describe as creative agency: from merely
referencing things (left hemisphere), towards describing things
(to help one's peers with recognizing, right hemisphere) by
creating a new word from two known constituents. (Say, berry-
bush). Of course, in proto languages the constituents were not
full words but much shorter phonemes, or just roots and extensions.
In German, the agglutinating is more obvious with our
compound words, whose first part is usually a 'determiner'.
So, determiner compounds might be an easier explanation for
the beginning of assemblages than predicatives. I don't know
if 'determining' is understood as a subcase of 'predicative’?

Mmmm. In Afrikaans a similar ‘compounding’ process takes place, where the
’rule’ is that one thing is signified by one word - however long, which can lead
to some absurd examples. Give me more examples of ‘determiner’ please …
I need to digest it, and am fascinated to explore whether it would function as a
good analytic bridge between indicative and predicative. I was not aware that
another category could be squeezed between indicative and predicative,
but of course that’s always possible!! Another layer of difference can surely (also) be another bridge, no?

Moreover, the process of taking two known, referencing,
left-brained, 'indicatives', and combining them to a brand new
language use that tries to 'show' something and to help
recognizing a (right-brained) experience, illustrates that these
two functions (referencing and showing) seem to be the basic
building blocks. I tried to express this earlier in my
comment to Keith Hamon here
and he agreed.

I will have to follow all these links … and get back to you.
I too realised that prepositions have a much stronger role to play than
I previously thought. See

4. Later, after the full predicatives, abstraction grew by
what Jenny mentioned in her post '18/9/16 JM': Halliday's
nominalization, when the predicate was shrunk again, into a

I think it might be much easier to follow your difficult
concepts of meta-semiotics and hyper-semiotics, when we can
ground our ideas about these on a full model of the simple
language explained in McGilchrist's terms.
That would be good. When I wrote the article on Fractured Narratives and Pop-up Diaspora (see researchgate) my editor asked me to take out the page on Agency, and suggested I should put it into a book - which unfortunately would only be read by about 5 people! I didn't think agency was that difficult!

Barthes, Roland (1977) Elements of Semiotics. London, Jonathan Cape
Cilliers, Paul (1998) Complexity and postmodernism: Understanding complex systems. London and New York, Routledge.
Cilliers, Paul (2005) Complexity, Deconstruction and Relativism. Theory, Culture and Society 22 (5) pp 256-267
Eco, U. (1978) Semiotics. Bloomington, Indiana University Press.
Foucault, Michel ( 1975) Discipline and Punish: the birth of the prison. London, Penguin.
Knorr-Cetina, Karin & Bruegger, Urs (2002) Global Microstructures: The Virtual Societies of Financial Markets. AJS, 107, January 2002. Pp. 905-950

Knorr-Cetina, Karin (2005) Complex Global Microstructures: The New Terrorist Societies. Pp. 213-234, Theory, Culture and Society. SAGE, London.

Latour, Bruno (2005) Reassembling the Social. Hampshire: Oxford University Press,
Snowden, Dave & Boone, Mary. (2007). A Leader’s Framework for Decision Making. Harvard Business Review.

Williams, Roy. (1983) The Impact of Informatics on Social Sign Systems. Pp 353-376 in B. Bain (ed.) The Socio-genesis of Language and Human Conduct. New York, Plenum.
Williams, Roy. (1993) Discourse and Text, pp343-363 In: Mass Media for the 90’s. A.S. de Beer (Ed), van Schaik Books, Pretoria, South Africa. Retrieved from:

Williams, Roy. (2005) Meta-semiotics and Practical Epistemology. Theory and Psychology, 15 (5): 719-734
Williams, Roy. (2006) Narratives of Knowledge and Intelligence: beyond the tacit and explicit. Journal of Knowledge Management 10 (4): 80-99.

Williams, Roy; Karousou, Regina; Mackness, Jenny (2011) Emergent Learning and Learning Ecologies in Web 2.0. IRRODL, Vancouver. Vol 12, No 3

Williams, Roy (2012) Affordances and the New Political Ecologies. pp.93-120, in New Directions in Terrorism Studies, Continuum, London
Williams, Roy; Mackness, Jenny; Gumtau, Simone (2012) Footprints of Emergence. IRRODL, Vancouver. Vol 13, No 4 (2012)

Williams, Roy (2013) Complexity and the New Ecologies of Knowledge. Data set published as supplementary resource: Retrieved 15-02-2015

Williams, Roy; Mackness, Jenny; Gumtau, Simone (2015) Synesthesia:
From Cross-Modal to Modality-Free Learning and Knowledge. Leonardo, 48: 1, pp. 48–54, Cambridge, Mass., MIT.

Zizek, Slovoj (2014) Trouble in Paradise. London, Penguin.

[i] Affordances, like signs, are similarly relational.

Just as the sign is to be found neither in the experience nor in the communicative interaction, but in the relationship between them, so too the affordance is to be found neither in the person (or agency) creating the affordance, nor in the ‘things’ – in the environment – that the person uses in a new way to create that affordance (Williams 2012).

[ii] See Susan Blackmore’s Memes and Temes TED Presentation. Retrieved 07.03.2015:

[iii] See Williams at al 2011 and 2012 on emergence and complex adaptive systems.