Tussling with seascape character assessment and assemblage theories

ABSTRACT Assemblage theory, based on the work of Manuel DeLanda, has gained in popularity amongst social scientists. However, a very different version of the theory has now been presented by Ian Buchanan, who claims that his version has much stronger roots in the philosophy of Deleuze and Guattari, from which the concept originates. A review of these two reveals the deep differences between them and their incompatibility, especially given the different ontologies upon which they are based. They can broadly be described respectively as systems-oriented and purpose-oriented approaches. These competing versions are explored further in the context of the practice of seascape character assessment. A twin analysis of a seascape character assessment exercise is carried out, which illustrates vividly the different understandings that each assemblage theory offers. However, this also suggests that each has the potential to help develop more progressive approaches to seascape practice. The choice of approach will depend on the aim of incorporating assemblage theory into this area of practice, whether to work within the rationale of, and enhance, existing assessment (DeLanda) or to interrogate the processes by which seascapes may serve particular purposes and then shape those processes (Buchanan).


Introduction
' … my version of the assemblage owes nothing to the work of Manuel DeLanda' (Buchanan, 2020, p. 16): so states Ian Buchanan in the introduction to his book Assemblage Theory and Method.This is a stark rebuttal of DeLanda, who is widely taken to be the leading voice of assemblage theory, claiming to have built a comprehensive model from the fragments of ideas about assemblage put forward by the French thinkers Gilles Deleuze and Felix Guattari in their seminal work A Thousand Plateaus (1988) and elsewhere.Buchanan qualifies his statement, but only to plunge the knife in more deeply, adding 'except perhaps as a useful reminder of why it is important to read Deleuze and Guattari carefully' (Buchanan, 2020, p 16)implying that DeLanda does not.DeLanda's two key books, including the one boldly entitled Assemblage Theory (2016), are effectively dismissed.
This presents a dilemma for social scientists interested in taking up assemblage theory in their own work: whether to accept DeLanda's version or reject this in favour of Buchanan's (assuming that these versions are, as implied by Buchanan's hostility, mutually incompatible).Assemblage theory has been widely welcomed and used in many fields, generally with reference to DeLanda's work; should we now pause and re-assess this dependence?This is particularly given that Buchanan is first and foremost a Deleuzoguattarian scholar (Buchanan, 2000) and may be offering us a take on the concept that draws more deeply on its origins.
Like many people attracted to assemblage theory, we come with our own specialist area of interest, in which we are searching for new, progressive approaches.Our area of interest is the practice of seascape character assessment (SCA) that seeks to evaluate the aesthetic qualities of coastal and marine areas; this is so that greater account can be taken of these qualities in planning processes, such as coastal development decisions.Our hunch is that assemblage theory can assist in developing current practice, by producing, for example, understandings of seascapes with a greater emphasis on relationality between their diverse dimensions which may in turn facilitate more finely attuned decisions.
However, we now have two competing versions of assemblage theory (and associated bodies of work) from which to draw.Even if initially attracted to the DeLandian model, we cannot ignore the alternative presented by Buchanan and his appeal to pay greater attention to the philosophical roots of the concept.But nor can we immediately dismiss DeLanda as Buchanan does; perhaps DeLanda's self-assured, philosophically selective and modified model is better geared to certain needs.
It is important to recognise that there have been numerous strands of thinking that take inspiration from the Deleuzoguattarian origins, such as actor-network theory and new materialism, as Buchanan himself mentions (2022, p4).We refer to some of this work below.However, we focus specifically on the polar opposites of DeLanda and Buchanan, and limit our detailed discussion to their main texts on the subject.
We therefore conduct something of an experiment, in which we explore these two options in relation to our topic.We do so by taking an individual seascape, as provided by SCA practice, interpreting it in the light of these opposing versions of assemblage theory and discussing their implications.This leads to some broad reflections on the possibilities offered by these options.We begin by reviewing the two models on offer and drawing out the considerable differences between them.

Splitting up assemblage theories
Dovey et al note that assemblage theory has been 'applied in a burgeoning variety of ways in urban studies' (2018, p. 265).The same is true more widely in human geography (Anderson et al., 2012;Anderson & McFarlane, 2011;Shults, 2022) with some concern for power relations (Allen, 2011;Legg, 2011).For example, McFarlane (2011) turns to assemblage as a way of emphasising the complexity of the socio-spatial processes that produce cities and give them identity and potential.Similarly, people in other disciplines have looked to assemblage as a way of breaking from conventional, deterministic conceptualisations, such as in archaeology, stressing the fusion of the immaterial and material in the development of meaning (Fowler, 2017;Joyce, 2020).There has also been some interest in assemblage thinking in coastal and marine studies.However, this has been largely dependent on the work of the geographers referred to above, with only superficial reference to the sources of the concept (Anderson, 2012;Bear, 2012;Fairbanks et al., 2018;Rossiter et al., 2015).It is also worth noting earlier marine-related work linked to actor-network theory, with which assemblage theory can be loosely associated (Callon, 1986).
In this work, DeLanda is frequently referred to as the architect of assemblage theory and is generally adopted unquestioningly (though with varying degrees of comprehension).For example, Dovey accepts DeLanda's own broad definition of assemblage as 'a whole "whose properties emerge from the interactions between parts"' (Dovey, 2010, p. 16;quoting DeLanda, 2006, p. 5).
The term assemblage, as explained by Buchanan (2015), is an imperfect English translation of the French agencement used by Deleuze and Guattari. For DeLanda, this 'refers to the action of matching or fitting together a set of components … , as well as to the result of such an action: an ensemble of parts that mesh together well ' (2016, p. 1).This suggests that his broad understanding of an assemblage is a system of parts that form a whole through various processes.Buchanan reckons that this misses the sense of purpose implied by the French term, which is better understood in English as an arrangement, as in a musical arrangement: a careful, deliberate composition, rather than a simple compilation of things (2015, p. 383; 2017, p. 458).
For DeLanda, the way in which assemblages form is through relations of exteriority.This is opposed to relations of interiority, by which things are connected by sharing inner properties, such as their genes.What matters instead are the capacities of things to affect and thus bring about connections.Assemblages are thus constituted of heterogeneous components that do not rely on shared properties, but come together to form a whole in which the components interact in complex ways.Interaction leads to new emergent properties, belonging to and defining the whole, which then act back on their parts, both constraining and enabling them (2016, p. 9).DeLanda claims that every social entity can be shown to emerge in this way (2006, p. 118).Relations of exteriority also allow parts to retain their autonomy such that they can be detached from one whole and be plugged into another (2016, p. 9).However, DeLanda does not explain what brings components together in the first place.This is one of Buchanan's criticisms; he reckons that DeLanda (and other versions of assemblage theory) overlook the concept of desire, which Buchanan describes as the fundamental driver of assemblage creation and the source of the purpose implicit in agencement (see below) (2020, p. 38).
For DeLanda, the heterogeneous components that constitute assemblages may be physical or social.This is illustrated by follow-on studies such as that of Durose et al, whose urban assemblages involve the 'nurturing and developing of heterogeneous resources such as relationships, knowledges and materials' (2022, p. 1).But these are not very clearly delineated.In contrast, Buchanan returns to Deleuze and Guattari's concept of stratification, which is largely passed over elsewhere.Stratification involves the grouping and consolidating of substances (Buchanan, 2000, p. 120).Strata comprise the geological, biological and alloplastic (the last of these being rephrased by Buchanan as techno-semiological, which refers to aspects of the human domain).These each operate in different ways, with increasing expressiveness in each type of strata, so that we humans, through the use of language and other signs, exceed the limits imposed by the earth's geology and our own biology, which are then transformed by us.This multidimensionality is key to Buchanan's understanding of assemblage, and is something that he says DeLandian variants overlook, as assemblages are 'reduced to mere apparatuses' (2020, p. 5).
Both versions of assemblage theory refer to material and expressive dimensions of the components that make up assemblages, the expressive being non-material aspects or forms that are more discursive in nature.DeLanda sees the two as opposite points on an axis, stressing the expressive role played by all components, such as body language (2006, p. 12).He also elevates the role played by 'specialised expressive entities such as genes and words ' (2006, p. 14).Buchanan refers to the material, but prefers the Deleuzian term content.His approach is less dualistic.He warns against seeing the material (such as actions, bodies and things) and expression (such as affects, words and ideas) as perfectly mapped onto each other: they are intertwined, but are separately formed; they resonate, but only to a certain extent (2015, p. 390).Moreover, the form of expression is performative, in that it does not just ascribe meaning to content, but shapes it and its wider role in an assemblage (2017, p. 472).
As mentioned above, Buchanan emphasises the crucial role of desire in assemblages (drawing on Deleuze and Guattari's work Anti-Oedipus (1983)).Desire is at the heart of assemblages; it moves over the intensities from which components originate, selects certain components and arranges them.It is the motivating, productive, organising force, to the point that 'The assemblage is an actual composition of desire ' (2005, p. 62, emphases removed).Desire is the basis of all behaviour, whether animal, human or more-than-human (2022, p. 56), and, at least as far as humans are concerned, operates primarily within the psychological realm.This is a challenge to DeLanda's emphasis on physical elements: 'assemblage was never intended to refer to ensembles of material things.It was always about the organization of desire' (p71).This clearly distinguishes Buchanan's assemblages from scientifically-conceptualised systems of the kind that DeLanda comes very close to.
Both versions refer to virtual and actual realities.For DeLanda, the virtual consists of the 'dispositions, tendencies and capacities ' (2016, p. 108) that set the conditions of the actual.The virtual has a mind-independent reality (DeLanda, 2002), but its ontological status is that of a possibility space that structures possible outcomes (DeLanda, 2005).The actual tends to be the product that arises from this, typically with material form.Buchanan complains that this is a reversal of things as set out (in a radical manner) by Deleuze and Guattari.For him, it is the actual elements, rather than the virtual, that are causative (Buchanan, 2015, p. 389).For instance, 'The actual is that which concerns the mind right now ' (2020, p. 59), and therefore has effect, whereas the virtual is inactive; it is what is felt rather than thought.Similarly, certain objects may be virtual, in that they are inert props that are not essential to the assemblage's working (2015, p. 389).Buchanan thus rejects DeLanda's description of the actual as the material expression of the virtual.
Both also refer to territorialisation, the process by which a territory is claimed and defended by an assemblage.For DeLanda, this may be a literal spatial process, such as the establishment of a neighbourhood, but can also be non-spatial, such as defining the membership of an organisation (2006, p. 13).Territorial stabilisation and homogenisation is achieved via coding introduced by the special expressive components of words and genes; this fixes the identity of the whole (2016, p. 22).In contrast, for Buchanan, territory is not primarily spatial or material, but is an act, a way of dealing with 'ever-present chaos' and gaining an order to life (2020, p. 85).He gives the simple example of whistling as a way of coping with anxiety (2020, p. 98).Territorialisation is therefore 'in our heads ' (2020, p. 96); material components may play a part in this, but territorialisation itself is their transformation into signs.
Both also refer to the countervailing force of deterritorialisation, which they conceptualise in their corresponding ways.For DeLanda, this means, for example, that spatial boundaries may be weakened by widerreaching communication technology (2006, p. 13).He considers the consequences here for his hierarchical system of assemblages (see below), such that deterritorialisation taking place at one level will have effects on other levels (2016, p. 83).Buchanan sees deterritorialisation very differently as a means of escape from a restrictive form of territorialisation and thus a potentially creative process.However, he warns against deterritorialising too much, as this can lead to a return to chaos; this is usually avoided by reterritorialisation of some kind (2020, p. 89).This illustrates that assemblages resist change and 'strive to persist in their being' (2017, p. 463).For Buchanan, there is a more fundamental, creative oscillation taking place between an organised transcendental plane and an unorganized immanent plane; an 'essential rhythm' between forces of organisation and disorganisation (2020, p. 52).
In addition, Buchanan refers to a principle of unity (also called the abstract machine) (2020, p. 121).This makes an assemblage work towards coherence.It also explains the limits of an assemblage, such as the rules that define a genre of music, effectively closing its formation; ironically, however, it may act to open it to change, such as by enabling innovation within a musical genre (2020, p. 45).
For DeLanda, a different end-point is in view.Assemblages form nested systems, in which lower-level assemblages act as components in higher-level ones.This leads to a worldview 'in which assemblages are everywhere, multiplying in every direction … And at the limit … we find the grand cosmic assemblage' (2016, p. 7, emphasis added).Moreover, for him, all assemblages have the same ontological status, regardless of scale; this is described as a flat ontology (2016, p. 13, 28).This gives the impression that assemblages are objective, observable entities comprehensively populating the universe.This is far from Buchanan, for whom assemblages are more elusive: 'the assemblage is a virtual entity with actual effects ' (2017, p. 473, emphasis added).They are also more scattered; he prefers to remain at the level of discussing how discrete assemblages work where it seems meaningful to do so, and this is precisely the purpose of the concept.'If everything is or must be an assemblage then the term loses precision, indeed it loses its analytic power altogether ' (2015, p. 391).
It follows that their approach to studying assemblages also differs.DeLanda proposes a quasi-scientific method, identifying and measuring the components of an assemblage by considering, firstly, variables (that is to say, the variable components that constitute an assemblage), and secondly, parameters (the values that those components range across).This allows him to speak of changing the settings of parameters to consider different scales of operation (2016, p. 83).(Here, as elsewhere, he self-consciously improves on Deleuze and Guattari by stating, for example, 'when the concept of assemblage is endowed by parameters … ' (2016, p. 108) a tendency that Buchanan derides.)Buchanan pays more attention to the purpose behind the mechanics and how consistency is achieved by the whole.For example, he stresses the need to investigate the interaction between the three types of strata, mapping the dependencies between them and the selectivity of components that is at work, seeing these methodological principles as a 'navigational aid' (2020, p. 30f).
Some of the differences between DeLanda and Buchanan's concepts of assemblage are illustrated by their varying treatment of examples from Deleuze and Guattari's war machine (1988).DeLanda refers to their manhorse-bow assemblage, a nomadic innovation for going to battle (op cit, p404).He sees this as an ideal example of combining heterogeneous elements: 'the personal, the biological, and the technological', (2016, p. 69), but does not describe their inner workings nor the relations between them.Instead, he turns immediately to their upscaling into the larger assemblage of an army.He then discusses the 'emergent capacities' that resulted from these larger assemblages, such as the ability of horse-mounted warriors to outflank an enemy (2016, p. 69f).
Buchanan turns instead to Deleuze and Guattari's example of the Roman infantry phalanx (op cit, p398f).He also refers to the combination of elements involved: the soldiers, weapons and armour that made the phalanx a powerful fighting force.However, he then points to the selection of possibilities that lies behind this, a key aspect of stratification.He describes the performance of the phalanx, whereby soldiers had to hold together and step into the place of any who fell.Buchanan then considers the values that were equally necessary to the army's effectiveness, such as courage, honour, and a conquering intent.There was therefore both content (the phalanx) and expression (militarism) involved, and a coming together of these two orders (2020, pp. 31-33).
We offer the following summaries of these two diverging and, as is now very clear, incompatible versions of assemblage theory.
For DeLanda, an assemblage is an entity composed of heterogeneous components that are connected by relations of exteriority (rather than by shared inner properties).These relations are based upon components' capacities to act, particularly when acting together.An assemblage actualises these capacities, which until then have been virtual (though they are no less real for this); it thus displays emergent properties.An assemblage is brought about and consolidated by two processes.Firstly, territorialisation, by which boundaries (which may be spatial or social) are established and the whole is homogenised.Secondly, coding, by which the identity of an assemblage is fixed, particularly through the expressive (as opposed to the material) dimension of components; specialised expressive components, especially language, are important here.There is therefore an upward and downward causality at work between components and an assemblage.Territorialisation and coding are parameters that can take on different degrees and lead to varying levels of stabilisation of assemblages, and can also act in reverse to destabilise them.Assemblages are nested, with smaller-scale assemblages becoming components of larger-scale ones, to the point that a single cosmic assemblage can be envisaged.All assemblages have the same ontological status, regardless of scale, and should be individually understood.In developing this concept, DeLanda draws extensively from mathematics and natural sciences, though his examples of assemblages tend to be socio-economic in nature, such as cohesive communities and hierarchies of economic activity.
Buchanan draws directly on Deleuze and Guattari for his understanding of assemblage, frequently countering the versions put forward by DeLanda and others.An assemblage is not a thing, but rather a device for explaining how certain things work, so attention should not be on trying to describe an assemblage but on using assemblage as an analytical tool, a way of mapping out how a particular phenomenon works.That said, an assemblage can be spoken of in fairly material terms.It is a multiplicity in which geological, biological and techno-semiological (certain human) components operate.There are orders of dependency between these and a power of selection at work ensuring the inclusion and exclusion of certain components.The components exhibit both content (such as bodies and actions) and expression (such as ideas and words); it is the resonance between these that produces an assemblage.The process as a whole is driven by desire, which acts as a productive and organising force.Desire draws on an underlying chaos of intensities and, in a process referred to as stratification, arranges them, setting them into systems of resonance.In this way, coherence is achieved, with the assemblage then sitting above its parts to maintain its own limits and unity.Desire thus actualises the virtual, in that what had previously only been tentatively sensed is made active.However, countervailing forces of undoing, or deterritorialization, are constantly at work, which in turn provoke reterritorialization; this reflects a fundamental, creative oscillation between an unorganized immanent plane and an organised transcendental plane.Buchanan's examples of assemblages tend to be drawn from specific psychological or social situations, such as a child's play or a case of policy development.

Seascape and assemblage
We turn now to the concept of seascape, and our perception that it may benefit from the perspectives that are offered by assemblage thinking.Seascape, in an aesthetic sense, 1 can be thought of as the coastal and marine variant of landscape.Like the concept of landscape, it deals with the values associated with distinct geographical areas with both natural and human features.There is a large body of work dedicated to understanding how landscape, in particular, is shaped and perceived (eg.Bell, 2012;Kühne, 2019).Here, however, we focus on a more policy-oriented approach, specifically in relation to seascape.In this context, an official definition of seascape is 'an area of sea, coastline and land, as perceived by people, whose character results from the actions and interactions of land with sea, by natural and/or human factors' (Natural England, 2012, p. 8).This notion has come to prominence through efforts to take it into account in spatial planning processes.In the UK especially, seascape character assessment (SCA) has developed as a way of evaluating seascapes along coastlines and out to sea (Hill et al., 2001;Tapper, 2008).SCA has now been formally built into UK marine planning processes, with the aim that seascape qualities should figure in decisions about the use of the coast and sea (Leyshon, 2018).
The government agency Natural England, working with the consultancy LUC, has developed a method for SCA, building on landscape practice (NE, 2012).This begins with the seascape wheel, which portrays the wide range of natural, cultural/social and perceptual/aesthetic features that combine to produce seascapes (Figure 1).These form the basis of generic seascape character types that typically occur in geographies such as English coastlines.Seascape character areas are then delineated within the planning area as a whole, each one made up of relevant seascape character types as shown on a map (eg, Figure 2).Each area has a textual and photographic   description of its features and may have a forward-looking assessment, such as a description of trends and pressures (eg Figure 4).
Assemblage theory, particularly the DeLandian version that is most commonly presented, has an immediate appeal to understanding seascape and taking forward SCA.It may, for example, help in understanding the way in which diverse elements combine to create a seascape whole.Dovey has interpreted urban landscapes in precisely this way (2010).It may also accommodate the continuity and fluidity that are distinctive of seascapes and the mobility of certain elements across seascape areas.The overview of Buchanan's version above suggests other possibilities too, such as understanding the underlying drive behind seascape designations and how they then work as coherent units.
We now explore these possibilities by considering how the two versions of assemblage theory might be used to interpret a particular seascape area.The area chosen is Croyde Bay, a small scenic bay in southwest England that is popular for beach tourism and surfing; it is enclosed by prominent headlands (including Baggy Point) and faces onto the Severn Estuary (Figure 3).An SCA was carried out here using the NE methodology as part of a wider assessment for the coastline (LUC, 2015) (Figure 4).We base our interpretations on this documentation and on our own observations carried out on site.
We refer below to theory 'DeLanda' and theory 'Buchanan', clearly referencing our authors but allowing for approximation and some wavering from how they themselves might express things.Page numbers below refer to the SCA report (LUC, 2015).

Croyde Bay according to theory 'DeLanda'
The material seascape of Croyde Bay, complemented by its SCA, is an assemblage.It is a definable entity, as evidenced by its well-defined and relatively harmonious nature.It consists of heterogeneous natural and human components as visually apparent and as described in the SCA, such as rocky platforms, a sandy beach, rounded headlands, a historic village core, tourist accommodation, people surfing and bathing, and seaward views.These are set out in the SCA as key characteristics (p136), reflecting the kind of seascape elements presented in the Seascape Wheel.They are not connected by any shared inner properties (indeed, they are too diverse in nature for this), but by their capacity to act together in forming a unit and generating an aesthetically pleasing whole.So the components listed above, along with others, combine to produce the sense of a bay characterised by properties that emerge from the interaction of these components and are thus only found at the level of the whole.In the SCA, these are described as special qualities, including such things as 'rich diversity of habitats … tourism hotspot … dynamic coastal landscape … enclosed nature …expansive views … ' (p138).Until actualised in the assemblage, such qualities remain hidden in a virtual realm.
A visual boundary is easy to set, landwards by the ridgeline and the headlands, and seawards by the horizon of the sea and distant land (though visibility shifts under different weather and light conditions).This territory is reinforced by the SCA which inscribes these boundaries on a map and homogenises the whole under the title of the seascape character area: Croyde Bay and Baggy Point (SCA 17, in the wider catalogue of SCAs).This expressive, linguistic coding is carried through in detail, with a description of the different components emphasising their interaction and contribution to the whole.There is reference, for example, to the 'sand dune system … supporting a nationally important mosaic of dune grassland, tall herbs and scrub', and 'high intensity use of the beach and dunes, particularly for surfing … ' (p136).
The degree of stabilisation or territorialisation of the assemblage can be calculated by considering, firstly, the variables that constitute it, and, secondly, the values (parameters) that they range across.For example, the variable of rockiness can have a range of values that are described in the SCA in quantitative terms: 'rocky wave-cut platforms of between 8 and 10 m high' (p136).A change in value may be such that a change in state also occurs, when, for example, rockiness increases so that a rocky foreshore is replaced by a cliff.The possible range of variables is also illustrated by social awareness of different features.This has a high value in relation to certain geological features, the coast 'being famous for large glacial erratics' (p136), but it has a low value in relation to certain habitats, as there is 'a general lack of awareness … about the marine habitats and species' (p139); the implication is that a higher level of awareness would further stabilise the seascape.Active forces of deterritorialisation are also at work, expressed in the SCA's forces for change, which are, for the most part, seen as negatively impacting on the seascape, such as scrub encroachment on maritime grassland and the extension and increasing permanence of holiday parks in formerly undeveloped areas (p139).These threaten the stability of the seascape as codified in the SCA.
Croyde Bay is not an isolated assemblage, but sits alongside others in a mapped series covering the wider coastline (Figure 2), connecting it with others further afield.Moreover, higher level seascapes can be envisaged, such as for the Severn Estuary as a whole, of which Croyde Bay and its neighbours would form components, and these in turn form components of higher levels still, such as for the UK as a whole, and, potentially, all the way to a global, or even cosmic, seascape assemblage.Conversely, each component of the Croyde Bay assemblage is a sub-assemblage, such as a rocky wave-cut platform or a village, each of which is then made up of sub-sub-assemblages, down to each rock or brick, or even more discrete elements.Each assemblage, at whatever level within the nested system, and of whatever seascape quality, has the same ontological importance.Also, individual components can belong to more than one assemblage and travel between them; for example, a headland may be a significant feature in more than one seascape, both horizontally and vertically through the system.A weather event may also figure throughout a nested system.Causality thus operates up, down and across the system, with a change in one component, such as an increase in shipping, or a new offshore wind array, potentially having knock-on effects throughout.

Croyde Bay according to theory 'Buchanan'
The material seascape of Croyde Bay, with or without its SCA, is not, in itself, an assemblage.What may be an assemblage is the means by which Croyde Bay is understood and works as a seascape and in which Croyde Bay's materiality and accompanying SCA play their part in producing a sense of seascape.Thinking of Croyde Bay as an assemblage is therefore a means for analysing it in this way and assumes that it is worthwhile to do so, perhaps because of its prominence in spatial planning and management processes.
The starting point can be, as for 'DeLanda', to recognise the multiplicity of components, though these can be categorised as geological, biological and techno-semiological.Taken literally, these categories can be mapped onto the key characteristics of the SCA, such as cliffs and rock platforms, grasslands and seabird colonies, and surfing and the maritime character.(Sensual and social aspects are therefore core components of this multiplicity.)More importantly, perhaps, is to consider the different ways of working that the three categories, or types of strata, have, and, in particular, the increasing expressiveness manifested in each of the types.This is most evident in the techno-semiological; for example, the 'iconic' Baggy Point coastguard pole 'has long been a favourite photo feature in people's visits to the headland' (p136), demonstrating the resonance between its material content and what it variously signifies to people, and the performative role then played by this form of expression (photographic moments) in the seascape as a whole.
The next step might be to consider how certain components are selected from the almost infinite range of possibilities; clearly not every grain of sand on the beach nor individual thought about the place is required.What is required are the components needed to compose a seascape.These are presented as the key characteristics of the SCA, expressed as headings and brief text, covering the range of elements set out in the Seascape Wheel and organised by the wheel's main headings.This selection and level of description is, no doubt, constrained by what is achievable under the terms of the consultancy contract for carrying out the SCA.An order of dependence between components is implied in the special qualities section that highlights features that are most sensitive to change, such as 'unique geological features and geodiversity' and the 'wild, untamed beauty of the large dunes' (p 138); these are thus elevated for attention in future planning and management decisions.During this stratification, other components are excluded; for example, modern 'residential development' (p 140) is briefly mentioned as a force for change, but is otherwise side-lined as a seascape feature despite its visual prominence (Figure 3); it is viewed negatively in the SCA, in contrast to the 'historic core of the village' (p 136), which is deserving of being a key characteristic.The SCA process therefore involves the deliberate selection, emphasising and organising of features that build up as coherent a picture as possible.
The desire lying behind this assemblage is implied in this consistency that the SCA tends towards: the purpose appears to be the production of a succinctly-named description which emphases the contained nature of the bay with its notable natural features, and which is popular as a seaside resort (p 135); an area that is distinct from surrounding seascapes, yet based on a commonly accepted approach and set of concepts (such as the seascape character types).The unformed intensities that feed this purpose can perhaps be detected in the textual material, such as the scientific knowledge and authority lying behind the description of 'soft Quaternary raised beach and periglacial deposits' (p 136) and the landscape tradition informing an appreciation of 'perceptual qualities varying throughout the day and season' (p 136).Virtual intimations of seascape are thus actualised in the performative language of the SCA.
Finally, the chaos of the immanent plane constantly presses in in the form of the SCA's forces for change, such as scrub encroachment on the upper grasslands.This provokes the counter-reaction of the transcendent plane in the form of management measures, such as the introduction of 'a traditional flock of Hebridean sheep to graze habitats' (p139) and keep scrub at bay.The SCA is thus a concretising of the seascape, a territory gained, stabilised and defended, with its clearly stated name and description, kept within its cellular, bounded limits, and its place established alongside others.

Considering the options
The twin analysis presented above is necessarily brief, but serves to draw out some of the stark differences between the two assemblage theories and the relatively little common ground between them.They part company early on, after a superficial sharing of assemblages involving a drawing together of heterogeneous components.The 'DeLanda' version turns to the emergent properties of the whole brought about by the latent capacities of the parts, and the interlocking, comprehensive array of assemblages of which any one assemblage forms a part.'Buchanan' focuses on the selection and organisation of parts into a coherent whole, driven by purposeful desire, in which content and expression resonate and forces of undoing are constantly rebutted in a creative dynamic.These might broadly be characterised respectively as systems-oriented and purpose-oriented approaches.As would be expected, both versions work within their own terms.However, viewing our seascape area through these two lenses suggests that each works better in some respects than others.
'DeLanda' works more easily within the existing logic and method of SCA, reinforcing many aspects of current practice.Its rationalistic, quasi-scientific approach, in which a comprehensive understanding is sought of the material and social components that make up an assemblage, aligns with the cataloguing of features that contribute to a given seascape.It reinforces the existing idea of a seascape area as constructed from a multiplicity of elements, accommodating well the heterogeneity of components as set out in the Seascape Wheel.The textual description of seascape features is supported by 'DeLanda''s emphasis on the expressive role of language within an assemblage.This is particularly helpful when referring to perceptual dimensions of those features, which relies strongly on interpretive language; the same can be said of describing a seascape as a whole.
'DeLanda' thus provides a framework for describing material-social-perceptual seascapes as assemblages and reinforces their representation by the SCA process.Other aspects of this approach also resonate with SCA.For example, it draws attention to the deterritorialising trends that work against seascape identity, often described as threats within SCA practice.Another strength of 'DeLanda' is that its vision of assemblages having systematic coverage and spanning multiple scales corresponds well with efforts to produce a spatially comprehensive system of seascape areas, with a nested sequence of smaller and larger seascape units.This opens the door for individual seascape components to be seen as belonging to, or travelling between, different areas, as allowed by 'DeLanda''s emphasis on exteriority.
'DeLanda' then offers additional features that are attractive from an SCA point of view.Crucially, it portrays components working together to produce a whole with properties that only manifest themselves at this higher level.This could help in attributing a distinctive overall identity to a seascape area that emerges from the symbiotic working of their parts, going beyond the rather atomised description of features that is typical of current SCA; naming and describing seascape assemblages in this way may distinguish them more clearly from one another.Another enhancement could be the recognition that components have varying importance in contributing to the whole, depending on their values, which may derive from either their material characteristics or how they are perceived.
More negatively, 'DeLanda' creates the difficulty of suggesting that every possible component and sub-component, along with their parameters, should be understood and measured at a level of detail that would defy analysis.
Turning to 'Buchanan', this version of assemblage theory represents more of a departure from SCA practice, though potentially offers valuable, new insights.It suggests a more probing approach to understanding seascape areas, which could then reshape how they are understood and managed.
'Buchanan' does connect immediately with some aspects of the existing rationale of SCA.For example, it draws attention to the potentially synergistic working of components, in the way suggested by the key characteristics set out in SCA, and how they build an overall seascape.But on most points, it forces new considerations.Firstly, in relation to seascape components, it stresses the resonance, albeit approximate, between the material content and social expression of key features in a way that is not clearly articulated at present.Along with this, it invites the categorisation of components into geological, biological and techno-semiological, with increasing means and degree of expressiveness through this sequence.Secondly, it draws attention to the selectivity at play within seascape characterisation, as it highlights that some components will be chosen for their significance, brought forward and prioritised in any given composition, whilst others will be deliberately excluded.This in turns raise questions about the reasons for these choices, and underlying power relations that may be at work.Thirdly, 'Buchanan' points to the overall purposefulness lying behind any seascape configuration.This may be related to the role a seascape plays, or could play, in planning and management decisions and wider policy processes.This suggests that SCA could focus, for example, on selected, key seascape areas that are the site of contention, face significant change or present new opportunities rather than be a geographically comprehensive system.Fourthly, this approach asks what motivating, unformed concepts and values are being drawn upon in seascape formation.This could open up consideration of the types of knowledge and the values informing the descriptions being made and the underlying assumptions shaping SCA processes.Fifthly, it produces a description of the ordering and structuring of a seascape composition that emphasises its overall character and identity more strongly than current practice.Along with this, it invites consideration of the forces of change, or destabilisation, at work and how these are then resisted and counteracted by a seascape assemblage.
'Buchanan' thus offers more radical, social insights to the processes operating in seascape production that could then be harnessed in developing more policy-focused representations of seascape.However, to carry this through to the practicalities of SCA, or wider seascape management, would require considerable reflection; it is possible, for example, that integrating a geographically selective focus and critical stance into existing policy and practice, as suggested by this approach, would expose the conclusions reached to continuing question and challenge.
One important point in relation to both approaches is that, initially at least, they presuppose existing seascape phenomena to analyse.Their implied methods are very different; 'DeLanda' lends itself to an objectivist approach (DeLanda, 2016, p. 108ff), whereas 'Buchanan' suggests an interpretivist approach (Buchanan, 2020, p. 122ff).But the two theories are each offering a tool for analysing things like seascapes, in the form of seascape character areas for example, that are already in operation rather than for devising them in the first place.
However, it is conceivable that adopting the stance of an assemblage theory can inform the whole process of seascape production and lead to representations of seascape that embody the emphases associated with the theory.It is with this in mind that we return to our opening motivation of searching for new approaches that are more rewarding than conventional ones, asking whether assemblage theory can assist in producing richer understandings of seascapes, which may then facilitate better planning and management decisions.Our conclusion is that each of the theories on offer can indeed contribute enhanced understandings, as set out in the discussion above.'DeLanda' would work more easily within the existing rationality of SCA, but expand its horizons, leading to a more multi-dimensional understanding of seascape structures.'Buchanan' draws attention to the formative processes that produce coherent seascapes that could be put to the service of seascape policy development.We are conscious here of Buchanan's own appeal for a return to the Deleuzoguattarian origins of assemblage theory; this is not simply in the interests of faithfulness to these origins, but out of a conviction that a faithful version of assemblage theory has an analytic power that can only be realised when it is taken fully on board (Buchanan, 2015).

Conclusions
The force of Buchanan's rebuttal of DeLanda's version of assemblage theory does not yet seem to have been fully appreciated.For example, Shults (2022) acknowledges Buchanan's complaint that the philosophical roots of the theory are being left behind, but describes him simply as being 'quite critical of DeLanda' (p 111) and proceeds to work regardless with DeLanda's model in line with his fellow human geographers.This does not do justice to Buchanan's passionate reclaiming and formulation of the theory.As we have shown in our review of the two versions of the theory, and their potential applicability to SCA, Buchannan is offering a fundamentally different understanding.This is based upon a Deleuzoguattarian ontology which he describes as, 'At its most elementary[,] … a dual system of an organized transcendental plane sitting on top of an unorganized immanent plane' (2020, p. 52); seascape comes into being in the transcendental plane, though perpetually oscillates with, and responds to, the unformed matter from which it derives.This is in stark contrast to DeLanda's flat ontology of 'differently scaled individual singularities ' (2006, p. 28) which suggests a much more static world; in relation to SCA, for example, this translates into interlocking and nested, equally-regarded seascape units that have comprehensive coverage.There is an irreconcilable divide between these worldviews which demands a choice when it comes to bringing this body of theory into applied fields, and a reasoning for the choice made.
It is clear that DeLanda's systems-oriented model has continuing appeal, and we are open to its usefulness in furthering the understanding and characterisation of seascape within wider endeavours of coastal and marine governance.However, we welcome the radical, purpose-oriented alternative presented by Buchanan.Choosing to work with one or the other will depend on the extent to which one wishes to expand the horizons of current practice and provoke consideration of the influences behind, and intricate workings of, the representation of seascapes.Similarly, the two versions of assemblage theory offer different possibilities to more theoretical work on land/seascape.
More generally, we suggest that wider uptake of assemblage thinking in the social sciences, as illustrated in our opening sections, would benefit from greater clarity about which of these variants, or indeed others, are being worked with.There is also scope for further elaboration of the two versions of assemblage theory than we have been able to present in this article.The implications of the clear divide between DeLanda and Buchanan could be profitably discussed in the growing number of fields of study and practice that have attempted to get to grips with assemblage thinking.As both versions claim Deleuzoguattarian credentials for their work, with different levels of justification, perhaps, we finish with words of Gilles Deleuze on the nature of theory: A theory is exactly like a box of tools … It must function.And not for itself.If no one uses it … then the theory is worthless or the moment is inappropriate.
(quoted in Bouchard, 1977, p. 208) Note 1. 'Seascape' has wide usage, from artistic representations of the sea to scientific descriptions of marine environments.Here, we use the term in a socio-cultural sense, referring to perceptions of coastal and marine areas (though different usages clearly inform each other).

Figure 3 .
Figure 3. Croyde Bay, North Devon (Photograph was taken by the authors).