Pebbles, peanuts, ponies and people

‘Agency’ seems to have become ‘a hammer’ for archaeology. ‘If I have a hammer, everything looks like a nail’ (Maslow Reference Maslow1966). Agency is now a quality that is attributed to entities as different as pebbles, peanuts, ponies and people (objects, plants, animals, humans) – and has attained an explanatory position of its own. While human agency is often regarded not as an individual quality, but as depending on social structures, contrastingly, animal agency is increasingly recognized, and even inanimate objects/things are argued to have agency ‘in themselves’, not only as ‘secondary’ or ‘distributed’ agency. This paper aims to discuss, and criticize, these uses of the concept ‘agency’ within archaeological theory, connect it to ideas regarding different ontologies, question the newness of the perspectives of symmetrical archaeology and materiality studies, and suggest other empirical and theoretical developments within archaeology.

Agency, as an explanatory concept, is a relatively new term within archaeology. Among the first to discuss it from a purely archaeological angle were Dobres and Robb (Reference Dobres, Robb, Dobres and Robb2000), and and it was elaborated by their co-authors. Dobres and Robb described ‘agency’ as a ‘notoriously labile concept’ (ibid., 8), used and developed within archaeology in very different manners (ibid., 3–14). The theoretical foundation of ‘agency’ in the humanistic and social sciences lies in the works of several scholars, particularly Marx (Reference Marx1959), Bourdieu (Reference Bourdieu1977, 70–76; Reference Bourdieu1990, 56–60), Giddens (Reference Giddens1979, 150; Reference Giddens1993), Gell (Reference Gell1998), Knappett (Reference Knappett2005), Knappet and Malafouris (Reference Knappett, Malafouris, Knappett and Malafouris2008), Malafouris (Reference Malafouris2013), Latour (Reference Latour1993; Reference Latour1999; Reference Latour2005), and others. Agency's history within archaeology is thoroughly presented by Johnson (Reference Johnson1989), Dornan (Reference Dornan2002), Gardner (Reference Gardner2004), Robb (Reference Robb2010) and Johannsen (Reference Johannsen, Johannsen, Jenssen and Juel Jensen2012).

Within archaeology, agency was first discussed and conceptualized around questions of human individual agency and creativity versus human agency as determined by social structures. Agency understood as the latter was proposed by Shanks and Tilley (Reference Shanks and Tilley1987), as constituted by social forces by Pauketat (Reference Pauketat2001, 74–79), as an interplay between social forces and individual enterprise by Bell (Reference Bell, Gardin and Pebbles1992, 39–40), as emanating from socially shared practices and meanings by Hodder (Reference Hodder1986), as emerging from the ‘lived life’ of individuals with ‘forward-looking intentionality and creativity’ by Hodder (Reference Hodder, Dobres and Robb2000, 23), and as ‘actions of individual social actors embedded within a broader socio-cultural and ecological setting’ by Joyce (Reference Joyce, Dobres and Robb2000, 71). With Robb (Reference Robb2010, 493) there is a turn to agency understood as a concept that has to be ‘historicized in specific contexts rather than used generically’, and in relational terms, as ‘a quality of the relationships in which humans act’ (ibid., 515). These relationships can be with humans as well as with contexts and things.

Agency is also a subject for other fields than philosophy and archaeology. Within psychology, Bandura defines agency as an ability to act consciously and intentionally, and as a prime force in human adaptation, development and change, on both collective evolutionary and individual levels, and human agency implies a moral agency through affective self-regulatory mechanisms. He claims there are continuous co-development systems and feedback systems between individual human agency, social structures and contexts/environments. All are intricately interwoven (Bandura Reference Bandura2006). This resembles Ingold's (Reference Ingold, Oyama, Griffiths and Gray2001, 27) ‘immanent intentionality of human beings’ engagement with their environment in the course of perception and action’. Bandura's agency is a systemic approach, like Robb’s, but with an impetus on humans as actants, with corresponding human ethical responsibilities. Things are not mentioned.

‘Things-in-themselves’ – agency or ‘animism’?

Symmetrical archaeology (Webmoor Reference Webmoor2007; Witmore Reference Witmore2007; Olsen Reference Olsen2010) focuses on ‘things-in-themselves’ and claims a non-a-priori distinction and division between human agency and material (non-human) agency. Everything (every thing) and everybody (every body) is a potential actant. When Bandura argued for co-developments and interactive feedback loops between society and individuals, symmetrical archaeology extends the range of interactions beyond the human realm by claiming that objects and people also interact and co-create each other.

Similar positions are proposed by others. Gell (Reference Gell1998) explored how people and objects ‘interact’ and how people experience objects as having agency and personality. He apparently struggled to reach a conclusive position, ending with allocating ‘primary agency’ to living entities and ‘secondary agency’ to objects. Latour (Reference Latour1993; Reference Latour1999; Reference Latour2005) developed the actor-network theory (ANT), in which he claims that humans and objects interact and co-act so intimately and symbiotically that division lines are practically fictional in the extended networks. Both people and objects are defined as ‘actors’, objects even as moral actors, co-responsible for consequences. Things have agency in that they delegate for humans. Hodder (Reference Hodder2012a, 95) argued for an entanglement and entrapment between people and things. But in addition to attributing to things a ‘secondary agency’, which Gell also distinguishes from the human ‘primary agency’, Hodder argues (Reference Hodder2012a, 216) that ‘in their objectness things also have primary agency’ that is not derived from human agency. ‘Things have primary agency in that they act in the world as a result of processes of material interaction, transformation and decay’. Symmetrical archaeology takes this position of things’ agency to the extreme.

Symmetrical archaeology has been seen as a further development of postprocessual thinking (Lucas Reference Lucas2012), even as a substratum of it (Hodder Reference Hodder and Hodder2012b). Contrastingly, according to Johannsen (Reference Johannsen, Johannsen, Jenssen and Juel Jensen2012, 310) symmetrical archaeology also grew out of a dissatisfaction with the postprocessual, constructivist position, a position that (according to symmetrical archaeology), first, represents an overly anthropocentric perspective where everything is described and explained from peoples’ positions and perspectives; second, focuses more on relations than on individual entities in the interactive processes; third, contextualizes explanations to the point of making people (as well as other agents/actants) into puppets on strings that are completely contextually determined; fourth, focuses on meaning (immaterial mental and cultural construction) as the primary explanatory force (causation); fifth, thereby, marginalizes the materiality of artefacts (‘things’) and people–artefact interactions by seeing them as simply expressions of culturally constituted negotiations of meanings, and therefore as secondary to processes of culture (Schiffer Reference Schiffer1999, 6; Olsen Reference Olsen2007); and finally, has an overly submissive attitude towards anthropology (Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012, 9). In sum, according to symmetrical archaeology, the postprocessual perspective has a tendency to reduce the materiality of archaeological findings and artefacts, as well as sites, almost to epiphenomena: only indicators and containers of human constructions, meanings and minds.

This confrontation with postprocessualism seems justified. Yet surely the postprocessual persistent insistence on the importance of meanings and social constructions, on contexts and on the relativism of idiosyncratic cultures, has fertilized the soil of archaeology and become schemata that most archaeological work now takes into consideration. However, postprocessualism's potential for philosophical relativism; multiverse interpretations; and multiple, relativistic and unstable social constructions, along with an insistence on ‘meanings’ almost to the point of oblivion to the material objects ‘in-themselves’, may have led archaeology away (perhaps even astray) from its identity as a science of material findings from the past.

Symmetrical archaeology claims that the move from processualism to postprocessualism (although certainly not a universal phenomenon in archaeology (Pearce Reference Pearce, Bintliff and Pearce2011)), implied a shift from a locus of explanation focusing mostly on nature-and-things, to a locus of explanation focusing mostly on people-and-society (Webmoor Reference Webmoor2007, 567, figure 2). Webmoor claimed that a symmetrical approach transcends both these different foci. He claimed that it thereby closes ‘the Great Divide’ in Western thinking between nature-and-things and people-and-society (Latour Reference Latour1993, 13 ff.), by having all the various articulations of people-and-things and nature-and-society as equivalent loci of explanation. He argued that they should not be kept distinct on analytical or explanatory levels: ‘This post-humanist repositioning de-centres humans as autonomous, independent beings in need of distinct explanatory concepts, making the non-(a)modernist recognition that things are just as much a part of being’ (Webmoor Reference Webmoor2007, 570, original emphasis).

Symmetrical archaeology's return to ‘things in themselves’ (Olsen Reference Olsen2007, 581) is thus not a return to the materiality of traditional archaeology: the counting, classification, and ‘catalogue-ization’ of artefacts, their materials, their manufacture and the interpretations of their presumed purposes. ‘Things in themselves’ is a dramatically different return to the material, with both the agency of things (inanimate agency) and the ‘thingliness’ or ‘materiality’ of humans accentuated (Boast Reference Boast1997; Webmoor Reference Webmoor2007; Witmore Reference Witmore2007; Knappett and Malafouris Reference Knappett, Malafouris, Knappett and Malafouris2008; Olsen Reference Olsen2010). Symmetrical archaeology, with its philosophical connections to post-humanism (e.g. Pickering Reference Pickering1995; Reference Pickering2001), distances itself from the descriptionism of processualism, from the constructionism of postprocessualism and from the anthropocentrism of both. It focuses rather on what ‘a new and unknown actor: the silent thing’ (Olsen Reference Olsen2003, 89) did, and does, to other things, to nature, to animals, to people, and to culture.

The inanimate agency ascribed to things is argued to not imply an anthropomorphic attribution of intentionality to objects which ‘animism’ holds.Footnote ¹ However, I find it strangely reminiscent of ‘animism’ when it is claimed that things

And that the goal of symmetrical archaeology, namely to level out differences between inanimate and animate beings, is ‘egalitarian’ (Olsen Reference Olsen2003, 88) and ‘democratic and inclusive’ (ibid., 98). Olsen reassures us, ‘Again, this is not a question of anthropomorphizing things, treating them as humans’, but then continues, ‘but rather of respecting their otherness and integrity’ (Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012, 203). Of this I must ask, ‘respecting’? Why not simply ‘recognizing’? Symmetrical archaeology seems to equalize (make symmetrical) the agency of things and people, the lifeless and the living, and apparently attributes them all with much the same kind of agency: an ability to act and inflict on one another.

To attribute agency to everything/every thing appears to represent a de-animation of living ‘things’: people and animals alike; and an attribution of an almost ‘animistic’ agency to ‘any-thing’. This symmetrical world of actant things and the ‘new and unknown actor: the silent thing’ (Olsen Reference Olsen2003, 89) is somewhat disturbing. It resembles the horrors of ‘The Thing’ of fiction and philosophy alike (Botting Reference Botting2012).

Symmetrical archaeology is not unique in decentring humans in the nexus of interplays between people, things, plants and animals. It is archaeological zeitgeist. Materiality studies and New Materialisms are relevant labels. In many respects this zeitgeist is an interesting shift from an anthropocentric to an omnicentric perspective: a levelling out of importance and meaning of humans and things in descriptions and interpretations, and the implementation of a post-humanistic change of punctuation of sequences where no entity is regarded as a primary cause, but as an element in multi-interactive causal sequences, networks and chains. This zeitgeist has many facets, and proponents with varying positions and profound disagreements (Ingold Reference Ingold2007; Knappett Reference Knappett2007; Witmore Reference Witmore2014; Ingold Reference Ingold2014). Even disagreements with themselves: ‘Symmetrical archaeology is not a claim to an undifferentiated world,’ wrote Witmore in 2007 (at 547), but in 2014 (at 206) this reservation is gone, as he claims that ‘air and soil, rain and sea, wooden doors and stone orthostrats, nitrogen-fixing bacteria and clovers, psycho-political commitments to Rome and Hadrian's Wall . . . all are things’.

This zeitgeist has positive potentials for increased understanding of human–animal–plant–thing interactions, and a return to a focus on things in archaeology without necessarily the explanatory appendices of ‘for what human purpose or meaning’ was this or that thing made, done or allowed to happen. These perspectives in archaeology have resulted in captivating presentations, such as studies of people's perceptions and experiences of their relations to ‘things’, in books like In defence of things (Olsen Reference Olsen2010), What things do (Verbeek Reference Verbeek2005), Entangled (Hodder Reference Hodder2012a), The transmission of affect (Brennan Reference Brennan2004), Archaeology. The discipline of things (Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012), Vibrant matter (Bennett Reference Bennett2010), Rethinking materiality (DeMarrais, Gosden and Renfrew Reference DeMarrais, Gosden and Renfrew2004) and particularly Handbook of material culture (Tilley et al. Reference Tilley, Keane, Küchler, Rowland and Spyer2006) and Excavating the mind (Johannsen, Jessen and Juel Jensen Reference Johannsen, Jenssen and Juel Jensen2012), and a multitude of papers. Many are tantalizing reading. However, I question the ‘newness’ of many of the New Materialists’, material studies’ and symmetrical archaeology's perspectives and findings.

‘Old wine in new bags?’

Although they are presented as new perspectives, arguing for the intimate, intricately interwoven and interdependent connections between entities (primarily humans and things, but also plants and animals), as well as the past and the present, I will say that they are not particularly new. The language and formulations, sometimes tantalizing and enchanting, but often impressively baroque, complex and opaque, represent perhaps a new genre in archaeological texts, contrasting the conservative clarity of more traditional texts. But the neologisms – enmeshment, entanglement, entrapment, embeddedness, fittingness, materiality, objectness, pragmatogony and others – do not in themselves guarantee academic theoretical ‘newness’. Nor does intricately complex language with almost impenetrable philosophical sophistication guarantee high intellectual standards or signal mind-blowing quantum leaps of paradigmatic shift (as Ingold (Reference Ingold2007, 2; Reference Ingold2014, 231–35) also spells out). I contend that many of the alleged new insights are ‘old wine in new bags’.

A few examples. It is commonly known that archaeology deals with old remains in various stages of (ongoing) decomposition. So Witmore writes (Reference Ingold2014, 213, original emphasis),

The newness of this information is not what strikes me.

Symmetrical archaeology and other materiality studies refer consistently to phenomenological philosophy. Its relevance for material-culture studies is thoroughly described by Thomas (Reference Thomas, Tilley, Keane, Küchler, Rowlands and Spyer2008, 43–72). Its contribution is particularly connected to clarifying the embodied engagements that take place between humans and things, to human being-in-the-world, and to perception. Thomas (Reference Thomas, Tilley, Keane, Küchler, Rowlands and Spyer2008, 48) refers to Merleau-Ponty (Reference Merleau-Ponty1962, 203) and explains his ideas about perception thus: ‘An embodied human being can see, and can move around, position itself in relation to things, and handle them. Sight, touch and movement provide quite particular ways of entering into relationships with things, and none of these can be achieved by a disembodied mind.’ Apart from the obvious mistake that ‘an embodied human being can see, and can move around’ (I know many ‘embodied human beings’ who cannot see or move around very much), and apart from the pathetic convoluted way of saying that seeing and moving are dependent on a body, what is new in this? Although published in 1962, even then people knew that most people could move and see.

Regarding enmeshments and networks and interactions between people, things and the world in general, other sciences also know this, and long have done. Psychological-developmental theory and research have since long recognized as axiomatic that nature and nurture (from food and family to cultural contexts and climate) interact inseparably (Bronfenbrenner Reference Bronfenbrenner1979; Feldman Reference Feldman2008). The same is recognized in medicine, genetics and epigenetics (Spector Reference Spector2012). The body and mind of the individual have long been known to be an ‘in-dividable’ unity (Borghi and Cimatti Reference Borghi and Cimatti2010). And in a wider perspective, that individuals are deeply integrated into and co-created with their geographical contexts, social ambiences, material creations and cultural contexts is considered basic and banal within psychology.

Biology has long recognized that the division lines between matter and organisms and between dead and alive are not that easily defined. There are continuous processes of life and death going on in every living entity, and lifeless materials are constantly incorporated into parts of living organisms and vice versa. Living creatures contain dead cells, death is not a point but a process, and viruses have been defined as both matter and organisms as they fulfil some, but not all, criteria of biological organisms. What appear to be stable individual entities are, in fact, unstable processually interwoven ‘entities’. This is axiomatic from atoms to galaxies.

To postulate that human beings’ production of, and interaction with, ‘things’ in the world have had transformative effects on those ‘things’ as well as on humans themselves should come as no surprise to archaeology or to evolutionary research (Deacon Reference Deacon1997). Two well-known examples can be mentioned. People started to cook, and the cooked proteins had profound effects on humans’ teeth and jaw muscles, which in turn affected the development of the human brain (Wrangham and Conklin-Brittain Reference Wrangham and Conklin-Brittain2003; Pennisi Reference Pennisi1999). And humans’ domestication of plants, wolves, sheep and cows had profound transformative effects on plants, wolves, sheep and cows, as well as on landscapes, human subsistence and material as well as immaterial culture, and finally ‘fed back’ to create changes in human bodies, brains and minds (Serpell Reference Serpell1995; Mlot Reference Mlot1997; PBS-Nature 2007). This interrelationship between mind and matter in the form of intricate co-development of brains, cognition and material culture has in recent years been solidly recognized (or rather ‘re-cognized’) in cognitive archaeology (Renfrew Reference Renfrew, Demarrais, Gosden and Renfrew2004; Malafouris Reference Malafouris2013). So to point to the difficult distinctions between, and deep entanglements of, dead matter and living organisms, processes and products, complex interactive mutual involvements, networks, and feedback loops between people, their surroundings and whatever people make or manipulate is to ‘reinvent the wheel’.

In the important text ‘Symmetrical archaeology. Excerpts of a manifesto’, Witmore (Reference Witmore2007, 556–57) makes a point of the phenomenon that structures made in the past are still parts of our lives today, that people walking down Oxford Street in London today are physically directed in their movements by the structures of the underlying Roman road (Via Trinobantina). He presents this as an example of the agency of a ‘thing’, the Roman road, and its continued presence in the present as an actant: ‘Here, achievements of the Roman period orient movement today, thereby impacting on thousands of twenty-first-century lives on a daily basis.’ An enchanting, almost poetic, way of describing how the past and present are inseparably blended. But it is common knowledge that history influences the present, that the past is present in various ways, in physical structures, traditions and memes. And as Witmore also says, many modern roads are constructed on top of ancient Roman roads. One could add: and they continue to connect cities also laid out by the Romans and still flourishing. His statement about Roman roads does not give any new information, but is old classical archaeological knowledge in a new wrapping, ‘old wine in a new bag’.

Agalmatophilia

Agalmatophilia is sexual attraction to, and sexual behaviour towards, statues, dolls and other human-shaped figures. I use it here as a metaphor for ‘love of things’. A common observation is that both lay and learned are criticized for irrationality if they anthropomorphize animals. Yet now, with ‘turn-to-things’ approaches, an anthropomorphizing of things seems accepted as rational and in vogue in archaeology. This is strange and contradictory.

Surely, objects, in both ancient and present times, are often attributed with anima (‘spirit’), identity, characteristics of a personal kind, and invested with human affection (Lindstrøm 2009a). For instance, in various cultures and epochs, weapons have been attributed with personalities and biographies (Pearce Reference Pearce2013; Gansum and Hansen Reference Gansum and Hansen2002, 16–17). And many of you who read this have spoken to your computer, and have, like Gell (Reference Gell1998, 18–19), experienced your car having personality and a will of its own, perhaps even suffered grief when dumping it. But still that does not mean, or prove with any scientific meaning, that objects actually and factually have ‘animacy’ or ‘agency’ resembling that of living beings. Therefore Gell wisely calls artefacts (things) ‘secondary agents’ with ‘secondary agency’ in contrast to the ‘primary agency’ in ‘primary intentional agents’ (humans):

But other positions attribute primary agency to material objects, things. That is ontologically and scientifically problematic.

Surely, things are important. Studies of human conceptions of things; attributions to things; relations to things; uses of things as markers of personal, social, political, philosophical, religious and existential meanings and as ways of constituting the ‘objective’ frames around peoples’ lives, thereby shaping human experience, activity and culture; and studies focusing on the fact that whatever humans do, both to themselves and to their surroundings, they do through/by/with things, are definitely highly relevant in archaeology. This is to ‘engage in the dialectics of people and things’, as Stanford University describe the aim of materiality studies in 2014 (https://web.stanford.edu/dept/anthropology/cgi-bin/web/?q = node/1097). But in fact this is largely what archaeology already is about.

We would not be the species Homo sapiens sapiens without our things. Human agency materializes, comes into being and evolves through/by/with things. Our lives are mixtures of minds and materials. Humans’ intricate involvements and attributions to things are relevant to archaeology, history, anthropology, sociology, psychology, philosophy, economy and possibly other fields, and hopefully in trans-scientific research, as envisioned by, among others, Ingold (Reference Ingold, Oyama, Griffiths and Gray2001, 255–79). Still, I claim that these studies tell us only something about ‘things-in-themselves’, but a lot about ‘humans-in-themselves’.

Humans love things. Humans’ emotional involvements with things are explicable. They are not examples of irrationality in humans. Contrastingly, they point to a deep-rooted human relationality (Bowlby Reference Bowlby1969; 1973; Robb Reference Robb2010, 494). This quality is typically human (but not exclusively so), and is an essential part of human nature and adaptation. A possibly particular human (or primate) twist of this tendency to relate, emotionally and socially, is the importance of face recognition (Tsao et al. Reference Tsao, Freiwald, Tootell and Livingstone2006; Tsao, Moeller and Freiwald Reference Tsao, Moeller and Freiwald2008). Psychoneurological research has shown that human beings are neurologically ‘wired’ (in the fusiform gyrus) to be particularly alert to faces and facial expressions. This explains why we tend to ‘see’ ‘faces’ in all sorts of objects and arbitrary stimulus complexes (Hadjikhani et al. Reference Hadjikhani, Kveraga, Naik and Ahlfors2009).

Likewise, that inanimate objects appear to have animacy, in that they may seem to act on their own, is because we are psychologically and neurologically impelled to be alert to movements, to interpret movements as ‘willed’, and to attribute agency and intentionality to anything that moves, particularly if it has something that looks like a face. This propensity made our ancestors alert to the presence of predators and other humans (Tinbergen Reference Tinbergen1951, 191–92; Atran and Norenzayan Reference Atran and Norenzayan2004).

We project onto our surroundings what and how we are. ‘Projection’ is the psychological term for attributing our own characteristics, emotions and motives unto others. This process is spontaneous, unconscious and strong, and is also directed towards animals and things. Therefore to feel phenomenologically that objects are animated and alive, and act on their own, is a common human phenomenon (Brown and Thouless Reference Brown and Thouless1965; Harvey Reference Harvey2005), and certainly not, as is often believed, only typical of children and hunter-gatherers. Amazingly, the ancient Greeks, famous for their philosophy, logic and early science, tended to perceive statues as animated (Schnapp Reference Schnapp, Renfrew and Zubrow1994), and some loved them even to the point of having sex with them (agalmatophilia) (Scobie and Taylor Reference Scobie and Taylor1975). To sum up: certain characteristics of human perception, projection and cognitive attribution processes, along with the strong human tendency to form relationships, make ‘animacy’ and ‘agency’ in inanimate objects very likely to be perceived and experienced – by hunter-gatherers, ancient Greeks, and modern university professors alike.

Mind and matter

But still, somehow people tend to know the difference. Although many of us speak to our computers or windowsill plants, most of us still recognize them as computers and plants, and would find it spooky if they spoke back.Footnote ² Still, Latour insists (Reference Latour2005, 63), ‘Objects too have agency’ and calls them ‘different types of actors which are able to transport the action further’ (ibid., 70). He writes,

And further down, ‘these implements, according to our definition, are actors, or more precisely, participants in the course of action’ (ibid., 71, original emphasis). I disagree. We say that ‘kettles boil water’, but in fact I/you/he/she/we/you/they boil(s) water in a kettle, cut(s) meat with a knife, hold(s) provisions in a basket, hit(s) a nail on the head (or on our own thumb, which seems the rule more than the exception) with a hammer, etc., etc. But the agency is human. We say that a ‘building impresses us’, but it is in fact we that have the impression. The building doesn't do anything to us in itself. These are metaphorical manners of speech, and should be recognized as such.

We have emotions towards and attachments to things. It is part of our common human nature and species characteristic. As several materiality studies have shown, things are part of our individual as well as group identities, contain our history and can be deeply significant. But things don't attach to us. Children know that dolls, although invested with affection, are dolls. Children know when they are playing, and are aware that they are playing. Sámi people know that sieidi (sacred boulders), although sacrificed to, and invested with emotions, history and symbolic significance, are boulders. Sámi, just as all human beings, have emotional relations to things. But most persons recognize that there are ontological differences, too. And the ‘too’ is the pivot point: it signifies the common ability to differentiate between phenomenology, the experience of things (their symbolic meanings, the bundle of projections, associations and memories connected to them), and ontology, the nature of things (their material constituents and properties).

Differentiations between ‘physicality’ (body, matter, exterior) and ‘interior’ (soul, psyche, mind) ‘seem to be universally present’, says Descola (Reference Descola2005, 140), who otherwise claims cultural diversity (ibid., 149). Likewise, Gell states (Reference Gell1998, 131), ‘It seems that ordinary human beings are “natural dualists”’. So apparently, despite our human propensity to animate things, despite our common manners of speech to talk about things in ‘animistic’ and anthropomorphizing terms, and despite the sometimes blurred overlaps of ontological qualities and apparent agency of things, people nevertheless recognize differences between living and non-living, exterior and interior, ‘body and soul’, mind and matter. In that our minds are right, and that matters.

Ontological differentiations matter. The fact that everything on this planet consists of atoms does not wipe out that there are quantum leaps between atoms and molecules, between any assembly of molecules and organic matter, and between blue-green bacteria (cyanobacteria) and mammals. Humans tend to perceive and experience these differences. (But please note: I am here not talking about the traditional scala naturae, the ladder where the status of a creature is defined according to the complexity of its structures. After all, spiders have eight eyes and eight legs, humans only two of each).

Irrespective of the interconnectedness and the difficulty, or perhaps impossibility, of defining absolute division lines between being and becoming, alive and dead, process and state, you and me, this and that, we still live in a human life-world where we, perhaps due to the inherent aspects of our species’ perceptual and mental apparatus, tend to recognize these phenomena as being different and separate, and also intricately interwoven, interacting and interdependent. It is usually quite simple for most mentally intact persons to perceive, experience, and understand both these aspects: the separateness and the unity, simultaneously, and without needing to attribute ‘agency’ (or any other kind of anima) to anything.

Yet the opposite standpoint has been voiced. Kant claimed that we can never know the world as it is in itself: ‘Das Ding an sich’, only as it appears to me (or us): ‘Das Ding für mich’ (Hartmann Reference Hartmann1871). Social constructivism and anthropology take this line of thought to the extreme by insisting on human perceptions and ontological categorizations as being relative, and often profoundly different between cultures (Nelson Reference Nelson1983; Burr Reference Burr1995; Viveiros de Castro Reference Viveiros de Castro1998; Ingold Reference Ingold2000, 25, 111–31; Reference Ingold2006). Contrastingly, Descola (Reference Descola2005) suggests a categorization of all ontologies into four basic types, thus reducing the idiosyncrasy somewhat. However, if differences in ontologies are really so profound and pervasive, one may wonder, how is it that people from various cultures can communicate and interact at all?

There seem to be enough common factors in ‘being human’ to make interaction possible. Language is one. The human perception and cognitive processing of the world are filtered and expressed through language. It is therefore not irrelevant to the question of ‘different ontologies’ that all languages share common structures to an amazing degree (Faarlund Reference Faarlund, Hessen, Lie and Stenseth2009, 229; Goettner-Abendroth Reference Goettner-Abendroth2012, 16), and that people can learn several languages. This points to essential similarities between humans and human cultures regarding perception and categorization of phenomena in the world. But paradoxically, on a detailed level, language can also be deceptive and convoluted, and difficult to decipher (Mumford and Anjum Reference Mumford and Anjum2011). This makes translations and interpretations difficult. As a consequence, this may interfere with our understanding of other people, and lead to ideas of so-called ‘different ontologies’.

Different ontologies?

As an argument for attributing agency to things, ethnographic and anthropological records of ‘different ontologies’ are often mentioned. It is noted that, in several hunter-gatherer and pastoralist cultures, specific animal species are regarded as another kind of ‘people’, or ‘kindred beings’ with minds, motives, personhood, individuality, intelligence and agency, transcending the species barrier (Myrstad Reference Myrstad1996; Ingold Reference Ingold2000, 69–72; Fowler Reference Fowler2004; Descola Reference Descola2005, 41; Willerslev Reference Willerslev2007; Alberti and Bray Reference Alberti and Bray2009; Fijn Reference Fijn2009). Interpreting this as a ‘different ontology’, Descola (Reference Descola2005) defines it as animism, implying that humans and animals are seen as having similar ‘souls’ (‘similar interiorities’), but different bodies (‘different physicalities’). Likewise, Viveiros de Castro (Reference Viveiros de Castro1998, 470, 478) calls this perspectivism: spiritual unity and corporeal diversity. In some cultures, also plants, stones, rivers and other phenomena of nature are reported to be regarded as animated, personal and possessing agency, shown, for instance, in food offerings to particular stones (Äikäs et al. Reference Âikäs, Puputti, Núnez, Aspi and Okkonen2009), and in greeting and hugging particular trees (Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012, 29–30).

Western negative stereotyping clichés regarding animism, primitivism, magic and shamanism have often coloured descriptions and interpretations of other peoples’ thinking (Willerslev Reference Willerslev2007, 75, 181–82). Attributions of ‘irrationality’ are often undercommunicated undercurrents. For example, it is regarded as correct, biologically speaking, when we say that humans are a kind of animal, but when ‘others’ say that animals are a kind of people, we call it ‘animism’.

Before jumping to the conclusion that there are ontologies that are completely different from and incompatible with the Western ontology, we need to scrutinize how these are verbally expressed. In particular we need to be aware of subtle nuances in research interviews with people from other language groups, and listen beyond our own cultural and linguistic categories (Kvale Reference Kvale2006; Willerslev Reference Willerslev2007, 185).

How ‘different’ is the alleged ‘different ontology’ of animism really? I am convinced that there are no differences between ethnic groups regarding rationality and irrationality. And cultures that are labelled ‘animistic’ and ‘shamanistic’ do recognize animals, stones and plants as being somewhat different and distinct from humans, and from each other. And lo and behold: in his detailed analyses of animism and shamanism, this is exactly what Willerslev (Reference Willerslev2007, 81) records when quoting his Yukaghir informant Vasili Shalugin: ‘Naturally, humans, elk, and bears think differently. They are different kinds of people.’ Now, when interviewed informants say that certain animal species are ‘another’ or ‘different kinds of people’, that clearly does not imply sameness. So we must scrutinize what the informants mean by ‘people’. It could simply be that, among them, it is recognized that these animal species share certain characteristics, abilities, habits and habitats with humans, that in these respects animals and people are similar, and can therefore both be labelled ‘people’. The informants’ use of the term ‘people’ in this connection may thus not be synonymous with ‘human beings’, but rather a category used for all species with certain characteristics in common. But this does not mean that humans and animals are so similar as to intermarry, as stated by animist shaman informants, and also noted by Descola (Reference Descola2005, 148). To transcend the species barrier is actually regarded as potentially dangerous, informants explained to Willerslev (Reference Willerslev2007, 89–118). After extensive interviewing and sensitive interpretation, Willerslev concludes that the animist and shamanist Yukaghirs in Siberia, both psychologically and linguistically, are deeply preoccupied with the differences between humans and non-humans, the dead and the living (ibid., 188, 190). Their world is absolutely not an undifferentiated fusion of ‘pebbles, peanuts, ponies and people’ with the same kind of agency.

Animist shamans are actually recorded as regarding their magic experiences and practices as explicable by, and non-conflicting with, natural science (Grøn Reference Grøn and Hardy2010).Footnote ³ These shamans expressed that what might appear to be ‘different ontologies’ are not mutually exclusive, as also Descola admits (Reference Descola2005, 147). Also the ontological typologies suggested by Ingold (Reference Ingold2000) and Pedersen (Reference Pedersen2001) are presented with similar reservations: the typologies are neither absolute nor mutually exclusive, and mixtures exist. Willerslev and Ulturgasheva (Reference Willerslev, Ulturgasheva, Brightman, Grotti and Ulturgasheva2012, 50) demonstrate that this lack of precise distinction between ontologies may be more the rule than the exception.

Projections and presumptions

Humans project and attribute agency and life onto things. As Kohring (Reference Kohring2014, 248) puts it, ‘it is a cognitive extension of self and experience to transfer anthropomorphic characteristics and actions onto physical objects’. But just as you and I speak to our things, but would be startled if they spoke back, I presume that Sámi who revere sieidi (sacred boulders) by sacrificing food to them would be surprised if the stones literally gaped and consumed the food. The Sámi girl who hugs a particular tree when returning to the summer camp after winter (Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012, 30) would be startled if the tree hugged her back.

Contrasting the idea of ‘different ontologies’, I would suggest that there are ‘degrees’, ‘levels’ and ‘variations’ in human experiences and beliefs that, in Western research, are labelled ‘animism’, an often grossly undifferentiated, erroneous and derogatory term.

1. First, as a species, humans share certain characteristics. Humans are neurologically ‘wired’ to experience objects in the surrounding world as animated and alive (Atran and Norenzayan Reference Atran and Norenzayan2004).Footnote ⁴ Therefore we project and attribute our characteristics onto beings and objects around us, presuming they are like us. This is shown, for instance, in folk stories and fairy tales worldwide, which are full of talking animals and animated, talking objects. Numerous television programmes with hidden cameras have shown that even modern, urban people are remarkably willing to respond to ‘talking objects’.

2. Second, on a collective level, cultures differ with regard to attribution of anima, agency, and personhood to non-human entities (animals in particular, but even humans for that matter: in earlier Europe and America women and slaves were presumed to lack or have inferior souls). But we should be very careful about interpreting the ethnographic and anthropologic reports, because linguistic meanings and nuances can be very subtle, and translation difficult.

3. Third, on an individual level, humans differ in the propensity to experience the non-human world as being animated and personal, and differ with regard to what that means for each person, regardless of collective cultural norms.

Consequently there is a variety of differences and similarities regarding ‘animism’, animism, and attributions of agency to humans, animals and objects that run criss-cross through our rough (and rude) distinction between ‘the West’ and ‘the Rest’.Footnote ⁵

Still, many will insist that there are cultures with ontologies that are profoundly different from that (or those) of ‘Western’ cultures (Hoskins Reference Hoskins, Tilley, Keane, Küchler, Rowlands and Spyer2006, 74–84; Ingold Reference Ingold2000; Reference Ingold, Oyama, Griffiths and Gray2001; Reference Ingold2006; Olsen Reference Olsen2010; Olsen et al. Reference Olsen, Shanks, Webmoor and Witmore2012, 29–35). This has given impetus to the attribution of ‘agency’ to things. Respect for ‘different’ ontologies seems implicitly connected to respect for the rights of indigenous peoples. However, this face-value acceptance could paradoxically imply disrespectful attitudes, lack of proper understanding, or both. Be that as it may. Respect for others does not preclude self-respect. For good or worse, this is the ‘catch-22’ of social constructivism: if all ontologies are equally ‘valid’, then the so-called ‘Western’, science-based one also is. So far, at least. It is possible, even likely, that ‘our’ scientific theorems and paradigms may be abandoned or changed, when proven wrong or inadequate through new research and evidence. (And let me emphasize: I use the term ‘science’ as a synonym for Wissenschaft (German) and vitenskap/vetenskap/videnskab (Scandinavian languages), a generic term for all scientific endeavours, encompassing both the natural and the humanistic and social sciences). These sciences seem reasonably well accepted around the world (‘Western’ medicine even embraced). Differences in culture and ‘ontology’ seem to represent no obstacles.

Knowledge gained in scientific empirical investigations, qualitative or quantitative, under the conditions of methodological stringency, stringent concept definitions, operationalizations, systematic quantitative statistics or systematic qualitative interpretational procedures, probability estimates, logically constructed interpretations, inter-rater validity or inter-interviewer agreement and (re)testable results, is the ‘gold standard’ of science. It is operationalized in the ideals of validity and reliability, concepts that are relevant in both quantitative and qualitative research. Scientific ‘truths’ (reasonably reliable and valid knowledge) are not established through democratic processes: neither by the decisions of the majority, nor by respect for other cultures or religious fundamentalists. To respect that some groups of people adhere to ‘different ontologies’ (if that really is a precise description) is not the same as ratifying them.

Peoples’ attributions and attachments to things and objects, and how people attribute agency to various entities, are legitimate and interesting research questions in themselves. But that is not the same as studying and establishing whether various entities actually have agency. It is here that symmetrical archaeology, New Materialisms, and materiality studies go astray.

Abolishing agency as a ‘one-size-fits-all’ concept

What consequences should the above considerations have for the concept of agency in archaeology? How can we define ‘agency’ and use it in a meaningful scientific way?

Logic is the basis of all sciences. So, first, logically speaking, if a term is applicable to everything (‘pebbles, peanuts, ponies and people’), it becomes nil, void of meaning, logically useless and void of explanatory power.

Second, the word ‘agency’ already has a rather well-established meaning, as Johannsen (Reference Johannsen, Johannsen, Jenssen and Juel Jensen2012, 339, original emphasis) defines it: ‘“agency” may be defined as the physical capacity to initiate actions on one's own behalf’ (actually very close to Bandura's definition). Therefore, to use the same word, ‘agency’, for a quality in, or of, entities that are vastly different ‘in-themselves’ (‘pebbles, peanuts, ponies and people’), and as a term for phenomena that arise out of these entities’ interplay, enmeshments, interconnections, networks and assemblages, and for a relational quality that is a temporary, contextual, situated and contingent occurrence, is bewildering and scientifically imprecise. Other terms than ‘agency’ could be coined and properly defined. ‘Secondary agency’, ‘object agency’, ‘distributed agency’ and ‘network agency’ may work. Neologisms can be very appropriate and justifiable. Academic concepts should be as precise and unequivocal as possible. It simplifies compiling, combining, comparing and communicating theories, hypotheses, methods and findings. It is vital for building theories. And, not least important, it facilitates meta-analyses. In short, it enhances scientific development.

Third, we should confront the materialism of some sections of post-humanism by being bold enough to differentiate between effects and acts, effectants and actants. If a rock (an inanimate material thing) falls down and crushes a house, it is not the rock's ‘agency’ that ‘did’ it. In fact, if anything should be attributed with causing or creating the effect, it is Earth itself, or rather Earth's gravity. The crush was an effect of gravity. The same holds for a tree (an inanimate living thing) looming over a house: if it falls down on a house on a stormy night, nothing about or within the tree itself ‘did’ it, although it is a living organism. The fall, and the subsequent crush, were effects of moving air and gravity. Also falling animals or people can have effects on whatever they land upon, but neither are those effects products of their agency.

Fourth, along with the differentiation between effect and act, I think we should reclaim the difference between the lifeless and the living in order to have a meaningful definition of agency. Cars were originally called ‘automobiles’. Despite the Roman proverb nomina sunt consequentia rerum (‘names are the essences of things’), to call a car ‘automobile’ (‘self-moving’) is objectively wrong. It does not move ‘by itself’. The driver makes it move. Computers and other electronic gadgets do nothing by themselves, not even when they ‘interact’ in complex ways, or ‘communicate’ with us. Everything they do is pre-programmed by people. It is people's agency, people's ‘distributed agency’, that is effected/performed through these things. Knappett's example (Reference Knappett2005, 25) of a pacemaker merging with the body of a person it is implanted in, and Latour's example (Reference Latour1999, 176–93) of what he described as a merged identity between person and thing – a man with a gun – and a distributed agency between them when shooting somebody, are both easily countered. Neither of these examples is a convincing example of human–thing merging or of inanimate agency. The pacemaker does not move by itself, its activity is enacted by humans through pre-programming; although implanted in a human body, is still easily discernible and extractable as a discrete object. And it is the man who shoots; and he does it with/through/by the gun, but he is the actant. The man could have chosen another weapon, the gun could not have chosen another man. Therein lies the difference. The pacemaker and the gun are vehicles for people’s choices, performance and agency. The ethical problems in attributing agency to objects are paramount. It is absurd to say that the gas was responsible or co-responsible for killing people in Nazi concentration camps. Only people were.

The lifeless are not actants, but can be effectants. Only living organisms can be actants, or have animacy, as also argued by Johannsen (Reference Johannsen, Johannsen, Jenssen and Juel Jensen2012, 338–40). But whereas Johannsen, correctly, points to vast inter-species differences, particularly differences between humans and other animals, I hold that we should first differentiate between living organisms with and without a central nervous system – non-cerebral life forms versus cerebral life forms.Footnote ⁶ If the above-mentioned tree's roots dig into the wall of the house, that is something such an organism is likely to do, but the ‘doing’ is not an act proper, ‘motivated’ in a central nervous system. Its ‘agency’ is at best a ‘secondary agency’ (Gell Reference Gell1998, 21), an effect of ongoing processes in the living cells of the plant, with no behavioural flexibility (intentionality, change or choice) involved. Alternatively, the plant's ‘agency’ could be called a reactive agency, in that processes within the plant are triggered by automatic reactions (re-actions) to chemical and physical stimuli. In contrast, if an insect starts to pull apart and eat the walls of the house, the insect indeed ‘does’ it, and is an actant proper. Its behaviour is an act. Although insect behaviour is considerably dominated by pre-programmed behaviours, an insect's behaviour still originates within an organism with a central nervous system that is complex enough to imply a certain behavioural flexibility (some capacity to make choices and changes, and to learn). The insect has an active agency. In vertebrates (fish, birds, mammals), and some invertebrates, this active agency with behavioural flexibility, the ability to act, and to act consciously and intentionally, becomes a main characteristic (Cambridge Declaration on Consciousness 2012Footnote ⁷). Cerebral vertebrates not only have effects on their surroundings and react to them, but have an active agency, and act on them.

Animal agency – and human

Our conceptions of (other) vertebrate and cerebral animals have varied, but apparently have always contained a certain fascination: what are they? How are they? Who are they? Peoples’ dependency on, and interaction, cooperation and cohabitation with, animals have triggered many categorizations and conceptualizations, ranging from regarding them as mindless things (Descartes Reference Descartes, Cottingham, Stoothoff and Murdoch1984) to regarding them as highly complex creatures, even divine beings (Graves Reference Graves1997; Ingold Reference Ingold1994; Gilhus Reference Gilhus2006, 95–100). The ancient Greeks and Romans, to whom we owe so much of our culture, had extremely ambivalent and complex relationships with animals (Gilhus Reference Gilhus2006; Calder Reference Calder2011; Lindstrøm Reference Lindstrøm2010), and people still do. Today, with the development of human–animal relations studies (Oma Reference Oma2010; Hill Reference Hill2013; Jennbert Reference Jennbert2011), questions regarding animal agency vibrate within archaeology.

The postulate of animal agency is often evaluated against human agency. Here psychology may contribute. Bandura (Reference Bandura2006, 164–65) aligns four core characteristics of human agency: intentionality (motivation, plans, strategies of performance), forethought (cognitive representations of anticipated outcomes), self-reactiveness (regulation and adaption of behaviour and plan execution), and self-reflectiveness (meta-cognitive reflection of plans and results, a self-aware evaluation of oneself). For comparison, I therefore suggest that animal agency be assessed against these four criteria.

The question then becomes, are there situations that require and encompass all these characteristics of agency, and thus can serve as touchstones for animal agency? Yes: situations that require complex (instrumental and observational) learning and situations that require complex, flexible problem solving. Such situations have already been extensively studied in animal subjects. The methodologies used have been scientific, stringent and rigorous, such as controlled experimental design and systematic observation. The results are astonishing. Here, only a few among an abundance of findings are mentioned: animals do manage complex learning and sequential problem solving with behavioural flexibility, forethought, self-reactiveness and reflectiveness (Midgley Reference Midgley1995; Bovet and Vauclair Reference Bovet and Vauclair2000; Bekoff, Allen and Burghardt Reference Bekoff, Allen and Burghardt2002; Griffin and Speck Reference Griffin and Speck2004; Auersperg, Kacelnik and von Bayern Reference Auersperg, Kacelnik and von Bayern2013), and have social learning (Bennett and Laland Reference Bennett and Laland2005). Animals make plans, and have intentionality and motives (Tolman Reference Tolman1948; Bugnyar and Kotrschal Reference Bugnyar and Kotrschal2002; Wilcox and Jackson Reference Wilcox, Jackson, Bekoff, Allen and Burghardt2002; Raby et al. Reference Raby, Alexis, Dickinson and Clayton2007). Some animals can, like humans, distinguish between intentional and accidental acts in others (Call and Tomasello Reference Call and Tomasello1998), a finding that indicates that ‘intention’ is a real phenomenon and a meaningful concept. More remotely related to agency, but not irrelevant to it, animals have emotions that are similar to ours and evoked for the same reasons (Seligman Reference Seligman1975; Overmier Reference Overmier1986; Midgley Reference Midgley1995; Panksepp Reference Panksepp1998a; Reference Panksepp2005a; Reference Panksepp2005b; Ganguli Reference Ganguli2006; Bekoff Reference Bekoff2007; Panksepp and Biven Reference Panksepp and Biven2010). Like us, animals live in social systems with complex hierarchies, behavioral codes and rules of conduct (Midgley Reference Midgley1995, 25, 66; Smuts Reference Smuts, Bekoff, Allen and Burghardt2002; Bearzi and Stanford Reference Bearzi and Stanford2008). Animals have individual personalities (Gosling and John Reference Gosling and John1999; Stamps Reference Stamps2007); many have self-awareness (Coy Reference Coy and Ingold1994; Panksepp Reference Panksepp1998a; Reference Panksepp1998b) and self-recognition in mirrors (Povinelli et al. Reference Povinelli, Rulf, Landau and Bierschwale1993; Plotnik, de Waal and Reiss Reference Plotnik, de Waal and Reiss2006; Prior, Schwarz and Güntürkün Reference Prior, Schwarz and Güntürkün2008), and can comprehend others’ perspectives (Kaminski Reference Kaminski2013; Clayton, Dally and Emery Reference Clayton, Dally and Emery2007). Animals have communication skills, and many even understand the communication of other species (including human language) (Bradbury and Vehrencamp Reference Bradbury and Vehrencamp2011). And this list of capacities is far from exhaustive (for more details see Lindstrøm Reference Lindstrøm2012 and Cambridge Declaration on Consciousness 2012). There are, of course, vast inter-species differences in agency as well as in other capacities, and vast individual intra-species differences. Nevertheless, the conclusion must be that animals (perhaps particularly vertebrate, cerebral animals) have actant agency proper.

So when hunter-gatherers, pastoralists and others tell ethnographers, anthropologists and archaeologists that certain animal species are ‘another’/‘different’/‘kindred’/‘kind of people’, this need not express a ‘different ontology’. If by ‘people’ is meant having cognitive and emotional faculties, social systems, communication, agency etc., these are not incorrect attributions to animals, ‘animism’ or anthropomorphism, but are realistic impressions based on observations of, and interactions and concrete experiences with, animals (Willerslev Reference Willerslev2007, 73–118). There is therefore no need to ‘accept’ their statements out of respect for non-Western ontologies. These peoples are simply right. Modern research has documented animal agency beyond any doubt. It is not a human construction, it is not dependent on our interaction with animals, and animals’ characteristics are irreducible to our representations of them (Reed Reference Reed and Ingold1988). We can perceive them reasonably correctly. (Here Kant was wrong.) When people perceive animals as animated and having agency, it is not, as with inanimate objects, because they appear to have faces and seem able to act; it is because they have (faces) and do (act). Thus, in the wording of symmetrical archaeology, animal agency can, indeed, be an example of ‘things-in-themselves’ and ‘their own inherent properties’ – or more precisely, animals-in-themselves. One must be permitted here to use an unpopular term: agency is part of animals’ nature.

Agency is also human nature. But not solely actant agency. We, other animals, plants and insects are also reactive actants. Our brains and bodies have numerous automatic reflexive aspects (Gardner and Gardner Reference Gardner and Gardner1998; Shackelford, Schmitt and Buss Reference Shackelford, Schmitt and Buss2005; Barrett Reference Barrett2010); and we have many unconscious and subconscious aspects (Pinker Reference Pinker1999, 3, 86–389; Ohman and Mineka Reference Ohman and Mineka2001; Ryan and Jethá Reference Ryan and Jethá2010) – perhaps more than we like to believe. Still, human agency is, to a considerable extent, conscious and intentional, an actant agency, an agency proper. It is a kind of agency that we have in common with other vertebrate cerebral animals. But human agency definitively has more far-reaching consequences than the agency of any other species. The level of cerebral complexity, cognitive complexity, capacity for abstract thought and behavioural and instrumental flexibility that characterizes Homo sapiens sapiens as a species is special. With it comes an ability to foresee complex consequences of behaviours and to make choices, and a capacity (potentially at least) for abstract moral thinking (Bandura Reference Bandura2006).

During the last decades humans have become able to predict even extreme long-term consequences of their behaviours (such as climate changes). That, combined with our agency to transform, on large scales, everything around us, implies a profound ethical responsibility for all life on Earth, a responsibility that is unique to humans,Footnote ⁸ a moral concern also expressed by Hodder (Reference Hodder2012a, 221). I hope we will never see ideas about distributed agency and shared responsibility used to exempt humans from their responsibility.Footnote ⁹ Human agency, individual as well as embedded in social systems, makes possible extremely complex interactions with, but also uses, manipulations and transformations of animate and inanimate objects, in order to create ‘things’ (artefacts, technologies), relations, surroundings and landscapes that may be purposeful for humans (although often catastrophic for other life forms). Actually, archaeology is the very science that has testified to this human agency, on both an individual and a collective social level, across times and cultures. Agency is indeed part of human nature, of humans-in-themselves.

‘Things in-themselves’ = the nature of things

To speak of ‘human nature’, or the ‘nature’ or ‘essence’ of anything, for that matter, has been banned from archaeological theoretical discourse for decades. ‘Essentialist’ has become an abusive term. This is odd, for two reasons: first, a conception of the species characteristics of humans, that there is a ‘human nature’ that is common to all humans, is so widely accepted as to be the basis for the Universal Declaration of Human Rights (UDHR): ‘All human beings . . . are endowed with reason and conscience . . .’ (Article 1), and for other international legislative measures expressing the principle that people should be treated equally. Equal rights rest on acknowledging a certain sameness regarding essential characteristics. The horrific consequences of abolishing this concept of a shared ‘human nature’ need not be spelled out: the Holocaust of the Second World War is pertinent.

Second, to abolish the concept of something's ‘nature’ or ‘essence’ is odd considering that archaeological practice increasingly employs, in very sophisticated ways, knowledge from scientific fields that do study the nature of things. Some examples: knowledge from geology, mineralogy and petrography is used to understand the properties of stones for tools, and the effects of volcanic eruptions (Sigurdsson, Cashdollar and Sparks Reference Sigurdsson, Cashdollar and Sparks1982; Goldberg and Macphail Reference Goldberg and Macphail2006; Zastawnya et al. Reference Zastawnya, Rauba-Bukowskab, Trąbskac and Trybalskad2013); botany provides understandings of the use and domestication of plants and agricultural developments; palynology augments studies of prehistoric climate and diets (Molloy and O’Connell Reference Molloy and O’Connell2004; Moe et al. Reference Moe, Fedele, Maude and Kvamme2007); zoology and ethology are used to understand animal characteristics (including the agency of animals), for analyses of hunting, domestication and other interactions with animals (see references above); biochemical DNA analyses are mandatory for studying human evolution (Ovchinnikov et al. Reference Ovchinnikov, Götherström, Romanova, Kharitonov, Lidén and Goodwin2000; Dalton Reference Dalton2010); dentistry, physiology and pathology are used to estimate age, sex, nutritional aspects, disease and causes of death from human bodily remains (Prag and Neave Reference Prag and Neave1997; Piontek and Vančata Reference Piontek and Vančata2012; Myszka, Piontek and Miłosz Reference Myszka, Piontek and Miłosz2012); and cognitive psychology and neuropsychology highlight the cognitive processes connected to the cultural changes occurring in the Upper Palaeolithic (Renfrew and Zubrow Reference Renfrew and Zubrow1994; Corballis and Lea Reference Corballis and Lea1999; Wynn Reference Wynn2002).

All this is already well-established archaeological practice. As a consequence, a return to ‘things-in-themselves’ and ‘materiality’ in archaeological theory should absolutely lead to (re)consideration of things’ nature, and its (re)inclusion in the analyses. These ‘natures’ are not ‘new things’, new concepts, or new fields to be discovered by theoretical archaeology or phenomenological philosophy. Sciences other than archaeology already have a good grip on the stuff that they investigate. The wheel does not have to be reinvented.

By continuing the already established tradition of archaeology as a multidisciplinary and interdisciplinary practice, it can continue to develop also as a transdisciplinary one. Multidisciplinarity is research on a common problem area, but with each discipline keeping within its own area. Interdisciplinarity is the integration of research and theory on a common research question. Transdisciplinarity synthesizes theories and data even more profoundly, and borders between disciplines can be dissolved. I agree with Ingold (Reference Ingold, Oyama, Griffiths and Gray2001, 276) that any ‘divisions . . . of inquiry must be relative rather than absolute, depending on what is selected as one's focus rather than on the a priori separation of substantive, externally bounded domains’. Both inter- and transdisciplinarity involve theoretical and interpretive challenges, and connect to challenging methodological integrations. Instances include methodological triangulation and various kinds of mixed-method research (Teddlie and Tashakkori Reference Teddlie and Tashakkori2003; Johnson and Onwuegbuzie Reference Johnson and Onwuegbuzie2004; Rothbauer Reference Rothbauer and Given2008). But the greatest challenge is perhaps that of attitude: the different paradigms and sciences must be acknowledged as being relevant for providing answers to different questions or to different aspects of the same questions, and as having theoretical paradigms and methodologies that are scientifically relevant and valid.

Here are four examples. First, burials of horses and dogs are found across cultures and periods. Their frequency represents quantitative data. Yet they are not explicable solely by the economic value and practical uses of horses, but are qualitatively interpretable by understanding the relationships between animals and humans (Argent Reference Argent2010; Jennbert Reference Jennbert2011; Lindstrøm Reference Lindstrøm2012). However, interpreting how such relationships could develop is dependent on including knowledge about inherent characteristics of both horses and humans, explored and documented quantitatively within ethology, psychology and neurology. Second, whether the Great Fresco in the Villa of the Mysteries in Pompeii shows cultic activities is highly disputed in classical archaeology. One piece of evidence would be exploration of whether the persons in the fresco wear cultic clothes. Such clothes are qualitatively described as having certain properties, such as colour combinations with symbolic significance. By using a quantitative test (chi square) it was shown that the colour combinations on clothes were hardly arbitrary but intentional, supporting the hypothesis of cultic activities. However, the symbolic meaning of the colours needed to be discussed qualitatively from science of religion (Lindstrøm 2009b; Reference Lindstrøm2014). Third, archaeothanatology has long been a transdisciplinary field. Analysis of human remains requires extensive collaboration from various disciplines, and provides information on various levels; for instance, forensic dentistry, pathology and osteology give quantitative information regarding the physical conditions of the person found, both in life and during the decomposition process. But it is vital, in addition to archaeology, to employ anthropology and science of religion for the qualitative analysis of the cultic and ritual aspects of mortuary practices. An understanding of natural and cultural contexts is dependent on knowledge in both the ‘hard’ sciences and humanistic sciences (Piontek and Vančata Reference Piontek and Vančata2012; Myszka, Piontek and Miłosz Reference Myszka, Piontek and Miłosz2012; Nilsson Stutz Reference Nilsson Stutz2010). Fourth, archaeoacoustics is a new transdisciplinary field. Sounds of various kinds are used in rites and in many cultic places, cross-culturally and from different epochs. Many cultic places have structures that magnify sounds. The acoustics must be measured by quantitative methods. Sounds could be used to evoke emotional reactions in listeners, as known from qualitative descriptions. However, how sounds may have been used to make people compliant and manipulable is explicable by a theory from social psychology developed through experiment and quantitative analysis (Dolinski and Szczucka Reference Dolinski and Szczucka2013; Lindstrøm and Zubrow Reference Lindstrøm, Zubrow and Eneix2014). Hopefully, these examples from varied areas and periods illustrate how different sciences, through theoretical and methodological integration, are already employed interactively to answer archaeological questions.

One could argue, against what I have said here about ontology, and against my advocacy for readmitting the ‘hard sciences’, that I may be ignorant of the critiques of science as a project that is epistemologically pure and free of values and potential political tendencies, and that sciences can be tainted with overt and subtle power relations. I am not. But somehow I have difficulty seeing that acknowledging ontological differences between objects/plants/animals/humans is politically, ethically and epistemologically problematic. In contrast, to blur them can be, as I have already argued regarding ethics and responsibility. True enough, sciences, both natural and humanistic-social, have been misused politically. Still (I may be terribly misinformed, but), I have never heard it argued that the ‘hard sciences’ – such as mathematics, glaciology, botany, astronomy, geology – are produced by particular structures of power or manipulated by race, wealth or gender. In contrast, these fields seem to be regarded as equally relevant and valued in all societies, from communist to fascist. In contrast, the ‘soft sciences’ – history, sociology, anthropology, theology – and the ‘mixed hard–soft’ sciences – like psychology and archaeology – seem much more prone to manipulation by politics and power. And it appears relevant and significant to this debate that ‘Western science’ (which is not only ‘Western’ in origin, but very much Arabic), is increasingly accepted globally. Under all political regimes and power relations, (‘Western-style’) universities are found and founded.

One admittance and some appeals

I have, in a perhaps provocative way, re-presented and represented the antithesis to the dominating theoretical paradigms of symmetrical archaeology, materiality studies, New Materialism and parts of postprocessualism. Nevertheless, I think it is deeply relevant within archaeology to do research on how materials effect (metaphorically ‘act and inflict on’) each other and disintegrate over time; on how people, animals and things (metaphorically) ‘interact’; and on how people perceived themselves in the world and perceived the world they lived in (their social, material, psychological, spiritual and contextual entanglements).

But, it is fully possible to recognize and do research and theoretical analysis on the enmeshments of time and periods, past and present, and of entities as different as ‘pebbles, peanuts, ponies and people’ within a more ‘classical’ ontology, one which differentiates between effects and acts, effectants and actants, and reserves ‘agency’ as a quality of cerebral life forms, animals and humans.

By turning the focus a bit away from overinvolvement with philosophy, symmetrical archaeology and materiality studies can have great potential, particularly in exploring perception and experience of practical interaction (workmanship/craftsmanship) with materials/materiality and materials-in-themselves (as also Ingold suggests (Reference Ingold2007; Reference Ingold2014). But to use the same concept ‘agency’ in all these connections is scientifically imprecise and creates theoretical complications. And why preclude recognition of knowledge and insight from other (perhaps even ‘hard(-er)’) sciences? They may provide some complementary, interesting and relevant ‘why-explanations’ to the ‘how-descriptions’. The grim ghosts of prejudice and ‘political correctness’ are haunting all exclusive paradigms.

Finally, it would be nice to see a development from thesis and antithesis towards synthesis in archaeological dialogues, theory and research. More theoretical and empirical inclusiveness, broader paradigmatic perspectives, more daring employment of divergent methodologies (mixed-method designs), more recognition of insights from other disciplines – in short, a further move towards inter- and transdisciplinarity. It might increase archaeology's ‘agency’.