Assemblage Theory Part 4: Actor-Network Theory and Latour

Rigid, reified categories limit our understanding of the world, and hence the need to focus on dynamic, ever-shifting networks and associations that connect events across space and time. Social realities are not out there to be discovered by either social actors or scientific endeavors; they are performed by multiple actors. (Siakwah, 2018)

Today I will take a  closer look at the Actor-Network-Theory since it also (!) uses assemblages, but I think in a slightly different way compared to Deleuze/Guattari.

I came across a GREAT website ( which was a great start for me to understand ANT a bit better.

As always, this page is intended for my notes and idea and therefore it will contain sources from non-academic texts/pages as well as many quotes!

Actor-Network Theory

Wikipedia’s webpage about ANT defines Actor-network theory (ANT) as “a theoretical and methodological approach [in] social theory where everything in the social and natural worlds exists in constantly shifting networks of relationships. It posits that nothing exists outside those relationships. All the factors involved in a social situation are on the same level, and thus there are no external social forces beyond what and how the network participants interact at present. Thus, objects, ideas, processes, and any other relevant factors are seen as just as important in creating social situations as humans. ANT holds that social forces do not exist in themselves, and therefore cannot be used to explain social phenomena. Instead, strictly empirical analysis should be undertaken to “describe” rather than “explain” social activity. […]

Developed by science and technology studies (STS) scholars Michel Callon and Bruno Latour, the sociologist John Law, and others, it can more technically be described as a “material-semiotic” method. This means that it maps relations that are simultaneously material (between things) and semiotic (between concepts). It assumes that many relations are both material and semiotic. Broadly speaking, ANT is a constructivist approach in that it avoids essentialist explanations of events or innovations (i.e. ANT explains a successful theory by understanding the combinations and interactions of elements that make it successful, rather than saying it is true and the others are false). Likewise, it is not a cohesive theory in itself. Rather, ANT functions as a strategy that assists people in being sensitive to terms and the often unexplored assumptions underlying them. […] It is distinguished from many other STS and sociological network theories for its distinct material-semiotic approach”.

So the questions that come up based on this Wikipedia article:

  1. “It posits that nothing exists outside those relationships”- What are the ontological/epistemological assumptions of ANT? Somewhere I read about the “flat ontology” connected to ANT, I think that it is connected to actors/actants being on the same level (but not 100% sure about that)
  2. “ANT holds that social forces do not exist in themselves, and therefore cannot be used to explain social phenomena. Instead, strictly empirical analysis should be undertaken to “describe” rather than “explain” social activity”- Is it just a way of DESCRIBING networks?
  3. What is a material-semiotic method?
  4. “It is not a cohesive theory in itself. Rather, ANT functions as a strategy that assists people in being sensitive to terms and the often unexplored assumptions underlying them”- I thought ANT was used to describe rather than analyze?
  5. It is distinguished from many other STS and sociological network theories for its distinct material-semiotic approach- What are the other sociological network theories? Do I know of them?
  6. Is ANT socially or technologically deterministic (or none?)?

So far the Wikipedia article has brought up more questions than it answered! However, I hope that I can answer them when I know about ANT.

Actor-Network Theory- Definition

I am not sure if it is possible to briefly define ANT, since there are different variations and also evolutionary versions of this..yeah what is it? Tradition, theory? Nevertheless, as Elder-Vass (2019) concludes, there are some elements and key concepts that are well-established in this tradition.

The main feature in ANT is its focus of actors and their effects on social processes. The actors in ANT can be either inanimate or humans, this perspective though si a radical notion as it attests agency to inanimate things.

“An actor can however only act in combination with other actors and in constellations that give the actor the possibility to act – this is because reality is assumed to be actively performed by various actors in a particular time and place. Thus inherent to ANT is a move away from the idea that technology impacts on humans as an external force, to the view that technology emerged from social interests (e.g. economic, professional) and that it thus has the potential to shape social interactions.”

Human ActorsInanimate Actant
Software developersLaptop

Cresswell et al. (2010) write that “ANT has its own epistemological and ontological position, in essence considering the world as consisting of networks. These networks can include humans, things, ideas, concepts – all of which are referred to as “actors” in the network. Tracing of associations or relationships between network components (or actors) is a key activity in ANT”. However, on some articles human actors are referred to as actants (e.g. Lewis and Westlung, 2015).

The authors further explain that “ANT assumes that the sum of non-social phenomena can account for something that is social as a result of constellations of human and non-human actors constituting the network. It follows then that the ANT approach is agnostic with respect to the debate which has divided many sociologists in that it neither asserts that everything is socially constructed (social constructionism) nor that everything is pre-existent (realism)”.

An assemblage is an assembling of elements and actants in a flat ontology, meaning treating the elements equally – in simple terms, a modern approach to the egg v. chicken debate.

This definitely answered my first question!

ANT seems to be prominent approach in geography and I found an article by Siakwah (2018) who points out that “ANT is also known as an enrolment theory, a process through which varied actors are recruited to form a network depending on their interests or a shared, emerging problem (Crawford, 2005)”. The author also introduces another central concept: translation, which is sometimes referred to as the sociology of translation. Translation in ANT, “is a process through which an actor recruits other actors into the network so that whatever the recruited actors do will help strengthen and further the interests of the actor who recruited them (Callon, 1981). The translation process encompasses creating connections and alliances and establishing communication among varied actors (Brown, 2002). It is the stronghold (or ‘centre of calculation’) position that enables an actor to control and utilize other actors’ activities to further his/her interest (Latour, 1988)”.

Networks in ANT

Networks play a central role in ANT, as ANT focuses on investigating and theorising “about how networks come into being, to trace what associations exist, how they move, how actors are enrolled into a network, how parts of a network form a whole network and how networks achieve temporary stability (or conversely why some new connections may form networks that are unstable)” (Cresswell et al., 2015).

As we already know, networks consist of actors, but there are also intermediaries and mediators that can form relationships between actors. According to Creswell et al. (2015), “the difference between the two is that the outputs of intermediaries can be easily predicted on the basis of their inputs (a black box). In mathematical terms, the assumption here is that X directly causes Y. Mediators, on the other hand, transform inputs into unpredictable outputs. This means that they can also transform actions, making something happen that is not necessarily related to what set it into motion. In mathematical terms, the effect of X on Y is in this case influenced by some other variable such as Z. ANT assumes that the social world consists of many mediators, which tend to be the focus of analysis as they impact on social outcomes in often unpredictable ways and very few intermediaries”. “A key task for the ANT researcher is to explore how local networks are ordered and re-configured over time.”

Another interesting point was that the “composition of networks tends to become particularly apparent when things in a system go wrong; conversely, these inter-connections tend to be hidden when things are working smoothly.” (Cresswell et al., 2015). This reminded me of the characteristics of infrastructures as described by Bowker and Star (2000), which become visible upon break-down.

My question here is, therefore:

What is the difference between an infrastructure (according to Bowker & Star) and a network (according to ANT/Latour)?

Actor-Network Theory vs Assemblage Theory/Thinking Part 1

Elder-Vass (2019) points out that one of the most fundamental claims of ANT is that “our world is composed of assemblages (also known as actants, actors, actor-networks, and articulations, to list only the ‘a’s).” This was initially quite confusing for me to read since I did not come across it in relation to Latour. Also, Elder-Vass accounts for that, as he explains that the concept originally derived from Deleuze. However, the author also points out that “the version advocated in ANT is distinctive – or at least, it seems to have narrowed down the range of meanings found in Deleuze. As in Deleuze, the word (derived from the French agencement) refers to something that falls somewhere between a thing and a process, with elements of both. For actor-network theorists, assemblages are open, transient, unique networks of influences or associations.” To support his claim, the author refers to a citation from Latour’s book “Reassembling the Social”; “an actor-network [= an assemblage] is what is made to act by a large star-shaped web of mediators flowing in and out of it. It is made to exist by its many ties: attachments are first, actors are second” (Latour, 2005, p. 217)”.

Elder-Vass added the assemblage in this sentence and is not found in the book. One time Latour actually mentions assemblage is on p. 2, where he talks about “the assemblages of nature” to “scrutinize what is ‘assembled’ under the umbrella of society”. However, he does not clarify it further I think.

My questions:

  1. I am wondering if assemblages according to Deleuze are NOT open, transient and unique? I am not 100% percent sure that these are characteristics of ANT, as outlined by Elder-Voos
  2. Does that mean that actor-networks are assemblages? Why or why not?

Elder-Vass (2019) further explain that assemblages do not correspond to everyday concepts of concepts or things, but that “assemblages are not bounded in the simple spatial way that ordinary things are. An assemblage is not a thing, but a coming together of influences”. According to Elder-Vass, this notion of assemblages in ANT is further illustrated in Latour’s discussion of the “work of the French scientist Louis Pasteur on the process of fermentation, which Pasteur attributed to the influence of yeasts. Pasteur, Latour tells us, “encountered a vague, cloudy, grey substance sitting meekly in the corner of his flasks and turned it into the splendid, well-defined, articulate yeast twirling magnificently across the ballroom of the Academy” (Latour, 1999b, p. 145). “Yeast”, in this sentence, is not simply another name for the same “vague, cloudy, grey substance” that had always existed; rather, it is a different assemblage, in which Pasteur’s own theories are among the elements or mediators that are bundled with or attached to the grey substance. Latour calls this articulation – yeast is an articulation of some material stuff and the ideas produced by science that relate to that stuff (and indeed other elements too, including perhaps the equipment required to support those ideas and the publications in which they are asserted)” (Elder-Vass, 2019).

Elder-Vass (2019) highlight another focus of ANT; the refusal of binary divides, such as “society and nature, human and non-human, and subject and object (Latour, 1993). Latour’s notion of yeast, for example, no longer locates it unambiguously in the categories of natural, non-human, and object, because for him the assemblage we call yeast includes textual, human, and subjective elements. One might question whether earlier thinkers are quite as dualistic as Latour suggests, but there is no doubt that he takes anti-dualism much further than most of his predecessors. Latour has strong views about the implications of this perspective for the research process. Research, he argues, must “follow the actors” (Latour, 2005, p. 68), or follow “associations” (Latour, Jensen, Venturini, Grauwin, & Boullier, 2012, p. 591) – the connections that make up assemblages”.

However, as the author points out, following the actors/associations raises challenges for research. Elder-Vass gives a couple of examples which I would support, and which I think also apply to assemblage theory according to Deleuze:

  1. How is a researcher to identify an actor/assemblage in the first place, when the boundaries of actors are so open and fluid?
  2. Do actors exist in forms that the researcher is to discover, or is the extent of the network of connections that make up an actor a construction of the researcher?
  3. Can boundaries ever be drawn between actors? How can we identify the constituent elements of actors when those constituents themselves are to be conceived of as assemblages rather than as conventional physical objects?
  4. How is the researcher to resist the collapse of analysis into a mélange of vague influences between unbounded networks? (Elder-Vass, 2019)

Even though some of the questions are addressed in research, Elder-Vass (2019) criticizes that “often these problems are resolved in a way that subverts the official ontology of assemblages. Without any systematic means of delineating assemblages, the elements that are put together to define an assemblage are just whatever mix of ordinary observable objects and people that the researcher thinks are relevant to the problem at hand – people, flasks, grey substances, and recording devices for example – rather than assemblages [actor-networks] as Latour defines them. The assemblage [actor-network] concept is then rolled out whenever this crude empiricism needs to be clothed with some philosophical respectability. In such cases ANT functions not as a solution to any of the classic issues raised by empirical social research but as a means of evading them”.*

* In the quote above I added the actor-network in the square brackets.

Harsh criticism! I am not sure if it is justified as I have not read enough articles and research that uses ANT, but I do think that the question he raises are valid and should be addressed because, at this point, this sounds like an IMPOSSIBLE task and approach to use in research. Well, it is possible, but how should research generate a GOOD analysis of an assemblage? Where are the boundaries, if assemblages are made of assemblages?

Actor-Network Theory vs Assemblage Theory/Thinking Part 2

The article by Müller (2015) focuses on the similarities and differences between ANT and AT, which was quite helpful for me, so here are the main points/quotes that helped me get it.

From the beginning on, the author highlights the significant similarities between ANT and AT: “One way to think of ANT is as an empirical sister-in-arms of the more philosophical assemblage thinking. Like assemblage thinking, it is interested in the provisional, socio-material ordering of entities beyond one universal principle: […]: ‘There is no social order. Rather, there are endless attempts at ordering’ (Law 1994, 101). […] The parallels between the concepts of the actor-network and the assemblage are significant.” While there are clear parallels between ANT and assemblage thinking, there are also notable differences and Müller (2015) presents what he considers the 3 major ones: agency, concrete concepts and politics.

  1. Agency

“ANT insists that agency is exclusively a mediated achievement, brought about through forging associations. There is nothing outside associations, and to become capable of action, entities need to form aggregates and find allies to produce an actor-network. Thus, what becomes political is a matter of what is made political through associations: ‘the political significance of materials is not a given; rather, it is a relational, a practical and a contingent achievement’ (Barry 2013b, 183).

With its focus on relations of exteriority, on the other hand, assemblage thinking posits that the component parts of an assemblage can have intrinsic qualities outside associations that impact on and shape the assemblage. It posits an open-ended set of capacities that is unpredictable and exceeds the properties of the component parts (B. Anderson et al. 2012, 179–181)”. As a consequence, seeing the world through associations, ANT has been criticised for being blind to what remains outside associations but may shape them nevertheless. (see criticism)

In the article written by Siakwah (2018) the author explains that “agency [according to ANT] is distributed across humans and non-humans (Latour,1999). However, things do not act by themselves, in a self-contained manner (Law, 1994); instead, the capacity for agency is a relational feature that emerges through interactions. ANT is premised on how things are formed and shaped through networks, associations and assemblages (Latour, 2005). Networks are made up of heterogeneous actors and their relationships (Johannesson and Bærenholdt, 2009). The relative heterogeneity of the networked actors ensures their resilience and durability (Latour, 1991, 1988; Doolin and Lowe, 2002)”.

2. Concrete Concepts

“Compared to assemblage thinking, ANT offers a more concrete conceptual and methodological apparatus that can be applied to empirical work. Terms such as ‘centre of calculation’, ‘oligopticon’, ‘black box’, ‘immutable mobiles’ and ‘translation’ or ‘overflows’ help to make sense of the formation of associations. The pioneers of ANT have delivered a string of analyses to illustrate these concepts. This makes ANT wieldier for empirical application.”

3. Politics

I am not really sure about this point. Müller (2015) states that “researchers working in the spirit of ANT have developed a much clearer notion of ANT’s relation to politics” and then continues to elaborate on ANT and politics, leading towards incentive to call for ontological politics or Dingpolitik, however, he does not really mention politics from an AT perspective, so I cannot really draw any conclusions here.

Criticism ANT:

“Critics have taken the approach to task for eschewing to think about how power differentials, for example, race, gender or class, impact on who or what is able or unable to form associations in the first place and thus for failing to acknowledge unequal power relationships. ANT also does not distinguish a priori between humans and materials, ignoring that humans are capable of intentions and pursue interests whereas things are not. With its task of following the associations that form networks, critics claim that ANT risks describing endless chains of associations without ever arriving at an explanation for the reasons and differences in network formation processes. In a similar vein, ANT discards social context, for example, cultural or historical factors, for explanation, unless it can be traced in the formation of concrete networks. In so doing, it also neglects to problematise the researcher and how his or her position is implicated in fashioning ANT accounts of certain phenomena” (Müller, 2015).

Interestingly, Cresswell et al. (2015) conclude that the aim of ANT is “to gain detailed insights into how social effects such as power come into being”. Following this, they present a parable, initially described by Law: “He describes how objects such as a big office, a computer and a phone can serve to create the manager in an organisation as the source of power. The manager studied in isolation (as a person or “naked ape” as Law calls him i.e. without objects), as opposed to as part of a network, is viewed as relatively powerless”. […]

ANT assumes that if any actor, irrespective of its position, is removed from or added to the network, as is the case if technology is introduced into an organisation, then the functioning of the whole network will be affected. However, networks are constantly evolving as social reality is assumed to be both complex and fluid.”

While I would agree with the second part of this quote, I am hesitant about the first part. I am not sure if ANT, in general, is looking into power. Quite the contrary, when I read some of the articles, this was one crucial point that critics mentioned: ANT does not account for power differentials and therefore fails to acknowledge unequal power relationships.

Bowker, Geoffrey C., and Susan Leigh Star. Sorting Things Out: Classification and Its Consequences. First paperback edition. Inside Technology. Cambridge, Massachusetts London, England: The MIT Press, 2000.

 Cresswell, Kathrin M, Allison Worth, and Aziz Sheikh. “Actor-Network Theory and Its Role in Understanding the Implementation of Information Technology Developments in Healthcare.” BMC Medical Informatics and Decision Making 10, no. 1 (December 2010): 67.

Elder-Vass, Dave. “Disassembling Actor-Network Theory.” Philosophy of the Social Sciences 45, no. 1 (January 1, 2015): 100–121.

Lewis, Seth C., and Oscar Westlund. “Actors, Actants, Audiences, and Activities in Cross-Media News Work.” Digital Journalism 3, no. 1 (January 2, 2015): 19–37.

Müller, Martin. “Assemblages and Actor-Networks: Rethinking Socio-Material Power, Politics and Space.” Geography Compass 9, no. 1 (2015): 27–41.

Siakwah, Pius. “Political Economy to Globalized Assemblages: Actor-Network Theory, Hydrocarbon Assemblages, and Problematizing the Resource Curse Thesis.” In Network Theory and Analysis, edited by Amber Jorgensen. Mathematics Research Developments. New York: Nova Science Publishers, 2018.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s