Artifacts and Attachment

Peter-Paul Verbeek, ‘Artifacts and Attachment: A Post-Script Philosophy of Mediation’, pp.125-146, in:
Harbers, H. (Ed.), 2005. Inside the politics of technology: agency and normativity in the co-production of technology and society. Amsterdam University Press, Amsterdam.


Within Technology Studies, the predominant vocabulary for understanding the role of artifacts in society is offered by actor-network theory. Bruno Latour, one of its major representatives, maintains that the social sciences’ exclusive focus on humans should be abandoned.

The so-called “principle of symmetry” is the most notable feature of Latour’s approach, entailing that humans and nonhuman entities should be studied symmetrically.


It will appear that Latour’s vocabulary is helpful in answering this question, but that it needs to be augmented in order to do full justice to the role of things in people’s everyday lives. I shall develop this augmentation by reinterpreting phenomenology, and by elaborating it literally into a post-script philosophy of technical mediation.


[Discussion of Eternally Yours industrial design]


But how should this materiality be conceptualised? How can one do justice to it when thinking about things – without recurring to a naturalistic or naive realist position which pretends to be able to get in touch with “things themselves”?

As pointed out in the introduction, Latour proposes treating people and things (“humans” and “nonhumans”) symmetrically. Actions are usually performed by compositions of humans and nonhumans. If I cycle to the university where I work, my travelling is performed jointly by my bicycle and I. There is neither any biking without me, nor without my bicycle. The nonhuman parts of such compositions should not be understood as passive and neutral instruments. They actively co-shape the action that is performed, that is, they co-act. Without a bicycle, my travel would be entirely different, and so would my contact with my surroundings.

Latour’s principle of symmetry is of great importance to developing a more material way of thinking about things. Not only does it show that things have a capacity to act, but also that they have this capacity by virtue of their materiality: their concrete “thingly” presence.

The specific way in which the bicycle, as a physical object, enables me to go to work organises my relation with the environment in a specific way. This “surplus” of what things do besides function can be described as mediation.


In On Technical Mediation Latour develops four concepts to help understand the mediating role of artifacts. He discerns translation as the first meaning of “mediation”. When artifacts mediate, they translate what Latour calls “programs of action”.

This emergence of hybrids can be indicated by the second term from Latour’s vocabulary of mediation: composition. Mediation consists of enabling the generation of new programs of action on the basis of the newly developing relations between the actants in question.

Mediation thus consists of mixing humans and nonhumans. This mixing work usually remains hidden, however. Most of the time it is “black-boxed”: the composite actant is taken for granted, making the “joint production of actors and artifacts entirely opaque” (Latour 1994, 36). This reversible black-boxing is the third meaning of “mediation”.


The fourth and last term in Latour’s vocabulary of mediation is delegation.

Such delegations, according to Latour, enable a remarkable combination of presence and absence. The past action of an absent actor – the designer of the speed bump, or the mayor who insisted that the bump be placed – exerts influence on people’s behaviour here and now.

These four aspects of mediation are closely connected. With regard to the speed bump, the authorities allied themselves to a piece of concrete (composition) and conveyed all that is needed for the realisation of their goal to the speed bump (delegation); after this, the bump can handle things by itself (black-boxing), because it changes the program of action of drivers from “driving slowly because of responsibility” into “driving slowly in order to save my shock absorbers” (translation).

Translation, composition, reversible black-boxing, and delegation each form an aspect of technical mediation that could not exist without the others.


If people are to be involved in the functioning of products, these products should delegate specific tasks and responsibilities to people. Delegations of this kind are underexposed in Latour’s analysis of technical mediation. In his examples, Latour focusses on delegations from humans to nonhumans. The speed bump, key weight, and door-spring aforementioned do what they do because people told them so; in other words, officials have bumps installed because they want people to drive slowly; housekeepers install door-springs to prevent draughts; hotel owners attach weights to their keys to stimulate guests to return them to reception when leaving the hotel.

This one-sided focus raises the suspicion of asymmetry. This suspicion is reinforced by Latour’s use of the concept of “inscription”. Scripts are supposed to be the products of “inscribing”, i.e., they are reducible to human activities.


Both Latour and Akrich see scripts as the result of “inscriptions” – and inscription is an asymmetrical concept, since only humans have the ability to inscribe.

Yet, Latour’s focus on delegation and inscription remains remarkable. If we are to understand the ways in which artifacts mediate, it does not matter all that much how they came to do so. What is important is that they play mediating roles, and the most relevant question to an analysis of technical mediation is how they do this.

For the understanding of technical mediation, the inscription processes and delegations from humans to nonhumans may remain black-boxed. Only the mediating role itself is relevant here, not its origins.


[…] Gomart and Hennion are right when they say that mediating artifacts should be approached as entities around which events occur, not as the outcomes of processes of interaction. This change of perspective creates the space necessary to see how artifacts actively co-shape the events around them, and to understand these events not only in terms of action but in terms of experiences as well. Delegations from things to people, not from people to things, form one way in which artifacts mediate what happens around them […]

The “script” concept remains biased toward action: scripts are defined as sets of “prescriptions” on how to act (Latour 1992, 232-233). And, more importantly, it erroneously suggests that mediation is a property of the artifacts themselves, not of the relationship between humans and artifacts.


An alternative interpretation of technical mediation can be developed, however, that does not localise mediation in the mediating artifacts themselves, but in the relationship between people and artifacts, or better such as in the “artifactually” mediated relation between humans and their environment.

In order to understand the “encounter” between humans and objects […] in a more detailed way than ANT does, a reinterpretation of phenomenology offers a suitable framework. In this reinterpretation, classical phenomenology is freed from the essentialist and romantic connotations that have become connected to it over the past century. Within the resulting “postphenomenological” perspective – to use a term of Don Ihde in a somewhat different way than he does himself – phenomenology is understood as analysing the relationships between humans and their world. Technical mediation should be localised precisely in these relationships.

During the first half of the 20th century, phenomenology was an influential philosophical movement. However, its influence has steadily waned over the past few decades. Phenomenology aroused the suspicion of being a romantic and essentialist approach. It was increasingly at odds with the rising contextualism in philosophy, brought about by the emphasis on both linguistics and postmodernism. Nevertheless, it is possible to formulate a phenomenological perspective that leaves these problematic connotations behind.

Like actor-network theory, phenomenology developed in opposition to the realism and positivism of the sciences. Seen from a present-day perspective, however, it ultimately did this in a highly problematic way. Against the claim of the sciences that they would reveal reality as it “truly” is, phenomenologists like Merleau-Ponty stated that scientifically interpreted reality was actually a derivative from a more fundamental one – that of the reality of everyday experience. Only on the basis of experiencing the meaningful, everyday world, Merleau-Ponty claimed, can the sciences build their abstractions. In opposition to the sciences, therefore, he saw phenomenology as a method for “describing” the world, not “analysing” it.

Classical phenomenologists saw it as their ultimate goal to “describe the world” (Merleau-Ponty) or to understand the “intuition of essences” (Husserl) or the “self-manifestation of being” (Heidegger), but in their attempts to reach these goals, they accomplished something else as well. In order to understand these “essences”, “world”, and “being”, they all developed analyses of the relationships between humans and reality. Husserl analysed this relationship in terms of consciousness, Merleau-Ponty in terms of perception, Heidegger in terms of being-in-the-world. It would be more modest and less problematic, therefore, to understand phenomenology as analysing the relationships between humans and their world.

Common to all classical phenomenological analyses of human-world relationships is that they contain a variant of Husserl’s concept of “intentionality”.

Husserl claimed that subjects and objects cannot be separated in this rigid way, but are instead always interrelated. The concept of “intentionality” indicates that humans are continually directed towards their world. People cannot help but be involved with their world as they are always experiencing it, and it is the only place where they can live their lives. In everyday life, subject and object are never separated, as realism and idealism suggest, but are always already interrelated.

Radically interpreted, this concept of intentionality is able to avoid the pitfalls of romanticism and essentialism mentioned above. Subject and object should not be seen as simply “related to each other”; they constitute each other in their interrelation.

In their mutual relationship, they co-shape one another. In each situation, humans are what they are on the basis of their relation to their world, and their world is what it is on the basis of their relation to it.


As later phenomenology did – including the later Husserl himself – this view of phenomenology replaces Husserl’s transcendental subject by an existential subject, and does not conceive the world as a collection of “objects”, but as a lifeworld. The world is the place where humans realise their existence, and which is continually experienced and interpreted by them.

What sets my reinterpretation of phenomenology apart from classical phenomenology, however, is the centrality of the notion of mutual constitution and, as will become clear below, the notion of the mediated character of this constitution.

The phenomenological perspective I defend here has the same symmetrical intentions as are present in Latour’s work in that it tries to overcome the “Grand Canyon” between subject and object by showing that both cannot exist separately. It even goes one step further, by stating that subject and object constitute each other.

Postphenomenology holds that realities come about in relations, as well as the humans that are related to these realities. Like actor-network theory, phenomenology stresses the contingency of reality, and the need for a “relational ontology” where reality is only given in the relations humans have with it.

The main difference between postphenomenology and actor-network theory is that postphenomenology is primarily interested in the relationships between people and the world, instead of the “constructions” that arise in and through them.


Postphenomenology does not bridge the “Grand Canyon” between subject and object by blurring the distinction between them – as ANT does, claiming that they are to be treated as semiotically equivalent entities – but by showing that they are intertwined, even at the level of their constitution. Postphenomenology aims to understand the “contact” between humans and world: the experiences and actions in which they co-shape each other.

Seen from the perspective of ANT, postphenomenology might appear to be a very limited approach, since it only studies very short networks. After all, the human-world relationships it analyses involve only two actants: a perceiving or acting human and an actant (human or nonhuman), which is experienced or interacted with. At best, three actants are involved: when technical mediation occurs, a third, nonhuman actant is added to the network.

But this simplicity is no shortcoming. The fact is that actor-network theorists could treat each of the three elements of the human-technology-world network as a black box, containing networks of any desired complexity. Postphenomenology does not deconstruct these entities, simply because it asks a different question than ANT does.

It is not primarily interested in the networks behind entities, but in the relationships humans can have with them – whether they are constructed or not. Postphenomenology studies these relationships in a more detailed way than ANT investigates the networked connections between actants.


The contact between humans and the world, therefore, has two modi: action and experience, aggregating into “ways of existing” (existentially) on the one hand, and “forms of interpretation” (hermeneutically) on the other.

The distinction between action and experience allows an expansion of Latour’s analysis of technical mediation. When an artifact is used, it co-shapes human-world relations by giving shape not only to people’s actions but also to people’s experiences.

This expanded understanding of mediation invites an expansion of the concept of “script” as well. Within Latour’s analysis, scripts are primarily related to action. Scripts concern the translation, inscription, and delegation of programs of action, whereas a concept is needed that indicates mediation in the broadest sense, a mediation of the interrelationship between humans and their world. Don Ihde’s concept of “technological intentionality” could serve this purpose, in a slightly adapted way.

With “technological intentionality”, Ihde indicates that technologies have “intentions” – they actively shape people’s relations with their world. A pen, for instance, asks for a completely different writing style than a typewriter and a word processor do.

As Latour does with “scripts”, Ihde implicitly localises this technological intentionality in the things themselves. From a postphenomenological perspective, however, it is more adequate to localise technological intentionality in the relationship between humans and their world.


This definition of technological intentionality implies that the mediating capacity of artifacts is no essential property of things themselves, but emerges from the interplay of things and their context.

Since technological intentionalities, unlike scripts, are not properties of artifacts themselves, but technologically mediated relationships via artifacts, it is not possible here to reduce artifacts to what was delegated to them by people.

Within a postphenomenological perspective, the reduction of nonhumans to humans or the other way round is simply not an option. Humans and the world are distinct and irreducible to each other, but they are nevertheless also inextricably intertwined and co-shape each other. Postphenomenology, in other words, does not abolish the distinction between humans and nonhumans, but shows their fundamental connectedness and interrelatedness.


Things mediate human-world relationships when they are used, and things-in-use are present in a peculiar way: they are present and absent simultaneously. When a technology is used, people’s attention is not directed at the technology itself, but at what they can do or experience by means of it.

When hammering a nail into the wall, people’s attention is not directed at the hammer, but at the nail; only when the head comes loose from the handle does the hammer ask for attention. Heidegger called these two modes of human-artifact relations “readiness-to-hand” (zuhandenheit – artifacts-in-use) and “presence-at-hand” (vorhandenheit – artifacts asking attention for themselves).


From the perspective of Heidegger’s distinction, the concept of “engaging technologies” might seem paradoxical at first sight. After all, how can a technology withdraw and at the same time ask for involvement and engagement with itself?

But, engaging technologies do exist. Artifacts that can ask for involvement do not necessarily have to be entirely present-at-hand.

Whereas a CD player only asks you to press a button, a piano demands an intense bodily engagement of the player.

An important dimension of this “engaging capacity” of artifacts is the skill that is needed to interact with them. Skill can be seen as the effort needed to “appropriate” artifacts.

The most important aspect of engagement with artifacts – which is also central in acquiring skill – is the necessity to interact with its machinery. The engagement a piano evokes is not comparable to the interaction with, for instance, a hammer. A hammer does not require active engagement with its machinery. For someone who has learned to use and “incorporate” a hammer, it becomes an extension of the body. A piano, by contrast, never becomes an extension of the body, even for the best piano players. It requires active and concentrated interaction with its keys and pedals, and therefore it never withdraws from a player’s intentionality relationship with his or her world.


Attachment comes about when artifacts invite engagement with themselves, and at the same time create scope for people to experience and interact with the world around them. They are somewhere in between “presence-at-hand”, by asking for engagement, and “readiness-to-hand”, by allowing humans to do something with them instead of only interacting with the artifacts themselves.

As the philosopher of technology Albert Borgmann shows, technologies tend to put their machinery in the background in order to allow people to enjoy the commodities they procure as quickly, easily, safely and ubiquitously as possible (Borgmann 1984, 41).


Understanding the attachment between humans and artifacts requires an attachment of actor-network theory to phenomenology.