Thinking with Deleuze and Guattari: An exploration of writing as assemblage

Abstract We need to keep experimenting with writing to meet the challenges of Deleuze and Guattari’s flattened ontology in the humanities. The paper reports on a small, experimental research project at a university in the north-west of England. The findings are written in an experimental mode, inspired by the Deleuze and Guattarian concept, ‘assemblage’. The experiment is theorised and assessed in a non-reductive way that offers future creative possibilities to other researchers. First, the paper presents a context for the subsequent experimental writing. Some current innovative writerly practice and some theoretical and methodological standpoints are reviewed. Next, this paper presents its theorisation of ‘assemblage’ with particular reference to Deleuze and Guattari’s use of the idea, ‘double articulation’. This approach supports and justifies the author’s schematisation of the textual assemblage into four areas: identity, work, territory and dissolving territory. The author explains how these ideas function within an experimental discursive text and illustrates their possible usage in the experimental text itself. Thus, this paper offers a theoretical justification, an explanation of and an assessment of experimental writing, in addition to the experimental text itself, all of which are of potential interest to researchers in the fields of education and philosophy.


Introduction
We need to keep experimenting with writing to meet the challenges of Deleuze and Guattari's flattened ontology in the humanities. This challenge can be summarised as marking a dissolution of the signifier-signified distinction. This paper presents a writing experiment that addresses the theoretical difficulties of flattened ontology and sketches a methodological approach for overcoming the wordworld dichotomy. This work is presented in a non-reductive way and does not invite reproduction through formularisation. Rather, this work is intended to theorise and demonstrate the workings of a possibility space as an avenue of inquiry for researchers working with the same theoretical-methodological dilemmas.
Deleuze and Guattari's ideas can be hard to handle within conventional discursive text. One explanation for this is that their flattened ontology ('plane of consistency' (Deleuze & Guattari, 1987, p. 73)) does not sit comfortably within the representational architecture of social science research papers. St. Pierre (2014) suggests that social science researchers influenced by Deleuze and Guattari may not These lengthy quotes from St. Pierre help me to set out this paper's theoretical and methodological priorities. First, following St. Pierre, I am treating my writing experiment as a form of thinking, rather than as a conduit for thought external to the text. This is significant because it takes the focus off the 'thinking human' and puts it onto the act of writing and the internal functioning of text itself. Second, this approach enables me to meet the challenge of Deleuze and Guattari's flattened ontology by thinking of text as assemblage. This concept will be thoroughly theorised in later sections, but can be summarised here as the idea that text possesses an internal dynamic through the co-functioning of parts. In other words, I will be concentrating on how the text asserts and maintains itself through its deployment of signs, with particular reference to Deleuze and Guattari's (1987) ideas, 'double articulation' and 'regimes of signs' . The key point is that signifiers moving at different intensities can become detached, or very nearly detached from their signifieds, while asserting their own agenda.
Thirdly, as St. Pierre's last quote shows, the text is ontologically equivalent to all other signs. My experimental text is therefore presented as its own enactment, not as a partial text in need of explanation. Nevertheless, in this paper I am conscious of the need to speak clearly to readers from the fields of educational research and philosophy of education. It is crucial to answer the 'so what?' question. I therefore include an explanation of what each component of the assemblage is intended to achieve, as well as an assessment of the experiment's success and usefulness. These additions are not regarded as 'the final word' or ontological solution of the experiment; they merely frame some ideas in a more explanatory way.
In the next section, I review the work of some researchers whose work has influenced my thinking and approach to this paper. Although presented in range of styles, genres and formats, all this work problematises the idea that signs can be deployed by 'humans' to represent 'the human' , in the sense of 'a subject who knows who she is, says what she means and means what she says' (Maclure, 2009, p. 104). In the section after that, I set out the paper's theoretical agenda; next, I move towards the experimental text, followed by a brief assessment and concluding remarks.

Writing after Deleuze and Guattari
Comtemporary researchers are continually developing new avenues of engagement with the creative and theoretical possibilities of the flattened ontology. First, I present some writers whose work has emphasised to me the creative possibilities of formal experimentation. Next, I review some work that engages with philosophical and methodological problematics. The following names are indicative of the diverse and still growing field and not a comprehensive review. Nevertheless, two significant points arise. Firstly, philosophical issues can be addressed through direct creative engagement and not only discursive reasoning. Secondly, there can be 'seepage' between different forms, styles and genres. There can be creative experimentation within a single genre, while creative and analytical writing can also be usefully combined. Both these points were influential in this paper. Wyatt and Gale (2018) write in the form of fluid creative exchange between voices in the form of a play script, in a style of 'making and undoing' they call 'clouding ' (p. 124). Their intention is to exploit the potentialities of writing in a 'non-totalising' format (p. 119). I found this work interesting for thinking differently about authorship, human situatedness and the suggestive power of artefacts. Also experimenting with the limits of genre, de Freitas (2017) uses speculative fiction to playfully consider how we might think beyond 'consciousness' to 'more-than-human research practices' (p. 125). De Freitas creates the fictional Laboratory of Speculative Sociology to help readers think about how our sensory environments are shaped by predictive technologies, and encourages us to build on these insights to shape the future of the social sciences. This was useful for thinking the limits of genre and the challenge of analysing ones own creativity.
In the fields of visual and performance art, O'Sullivan (2016) draws on Deleuze and Guattarian concepts to support his idea of 'fictioning' . O'Sullivan critiques the 'archival turn' in contemporary art practice that seeks novelty through reorganisation rather than true creativity (p. 81). O'Sullivan writes (in a way suggesting the influence of Deleuze and Guatarri's 'regime of signs') that a more radical approach would see artists coding independent worlds, but outside the dominant codes, between registers, splicing current systems of signs with new experimentations (p. 84). This was useful for my thinking around different semiotics (image-performance-text) from which new theorisations can arise.
Deleuze and Guattari's work creates problems for data coding and analysis that are particularly relevant to this paper. Researchers such as Cole (2012) Jackson and Mazzei (2012) and Maclure (2013) problematise empirical data that addresses the 'human' behind the text. Cole argues that Deleuze and Guattari's immanent materialism enables the use of data as a 'form of becoming … full of sometimes contradictory and elusive phenomena, such as unhuman becomings or affective assemblages' (2012, p. 14). Cole points to the need to move beyond subject-centred accounts of teacher learning. It can be more productive, Cole suggests, to think of the teacher as 'becoming-cyborg' (p. 15) as they increasingly map their teaching practices onto classroom technologies like interactive whiteboards and smart lecterns. These interactions redefine knowledge as an ability to function in technologically defined spaces (p. 15). Jackson and Mazzei (2012) challenge the conventional research signifiers of human presence (data, voice, narrative and meaning-making) that result in 'data' and lead onto 'analysis' (intro., p. viii). Analysis, understood as reduction of plurality to essential categories, effaces the emergent and multiple in the human. Jackson and Mazzei (2012) talk about 'plugging in' (p. 1) with theory, but not in discrete concepts. Rather, they favour 'schematic cues' from Deleuzian theory (p. 87) that involve 'thinking with' and towards their research project. Thus, conceptualisations do not occur in the abstract but in process, marking the space between theory and empirical work without precise boundaries. This paper is particularly influenced by Maggie Maclure's (2013) work on data coding in education research. Maclure is referring to the practice of fixing data into categories. She says 'coding assumes, and imposes, an 'arborescent' or tree-like logic of hierarchical, fixed relations among discrete entities' (p. 168). That is, coding subsumes data under the ruling idea and fixes down its internal relations. Coding as 'categorisation' , 'does not allow that things might (will) deviate and divide from themselves and form something new. It cannot cope with difference in itself-as movement, change and emergence' (p. 169). As Maclure suggests, one avenue of response is to keep alive ones capacity to wonder about the data, 'to allow something other, singular, quick and ineffable to irrupt in the space of analysis' (p. 164).
I will now sum up the main points about writing. Writing can be seen as way of 'thinking with' a non-humanist ontology. It can be seen as a way of making difficult (impossible) thinking actual, in an experimental rather than representational medium. These ideas are offered tentatively rather than dogmatically. Researchers have pushed the boundaries of writing after Deleuze and Guattari towards playful, creative and critical forms that trouble the idea of human presence. It is particularly important for this paper to trouble the idea of data as an empirical phenomenon, and to question the role of analysis in educational research, understood as fixing down the identity of data through coding.
In the next section, I will trace a Deleuze and Guattarian ontology of language towards a theory of linguistic assemblage.

Double articulation
From linguist Louis Hjelmslev, Deleuze and Guattari (1987) borrow the idea of 'double articulation' . Here, the binary schema signifier-signified is replaced with 'expression'-'content' , while each of these terms is 'itself articulated or composed of both form and substance: the first articulation correlates form and substance of content; the second correlates form and substance of expression' (Holland, 2013, p. 57). An illustration of how the double articulation functions in practice, is found in Deleuze's work on Foucault. In this passage, Deleuze explains the double articulation (expression-content) operating in the penal system. The content has both form and substance: for example, the form is prison and the substance is those that are locked up, the prisoners … The expression also has a form and a substance: for example, the form is penal law and the substance is 'delinquency' in so far as it is the object of statements. Just as penal law as a form of expression defines a field of sayability (the statements of delinquency), so prison as a form of content defines a place of visibility ('panopticism' , that is to say, a place where at any moment one can see everything without being seen). (Quoted from De Landa, 2015, p. 38)

De Landa (2015) then offers further examples of how double articulation (sayability/visibility) can apply in practice:
Deleuze is here distinguishing the two articulations roughly along the lines of the discursive (coding) and the non-discursive (territorialisation). Non-discursive practices of visual surveillance and monitoring, performed in buildings specifically designed to facilitate their routine execution, sort the raw materials (human bodies) into criminal, medical or pedagogical categories; and discursive practices, like those of the criminologists, doctors, or teachers who produce a variety of conceptual categories, consolidate those sorted human materials, giving prisons, hospitals and schools a more stable form and identity. (2015, pp. 38-39) There are three key points for this paper. Firstly, signs and bodies are not separate from one another, but affect and disturb one another-bodies elicit coding practices, coding alters bodies. Secondly, language can detach itself from formalisations of content (Holland, 2013, p. 82) and can be an assemblage in its own right (De Landa, 2006, p. 15). Here, the emphasis is on the expressive power of mixed semiotics to overwrite the actual as code. For example, take a social event. People post online images of the event, while the event is still happening. These images elicit further responses from participants also during the event. Here, the actual event (conversations, actions, groupings, words) reconfigures itself to mimic its online counterpart. As the behaviour changes, the virtual event overcodes the actual event. Thirdly, a linguistic assemblage asserts its own work and own identity, for example, when an online social event overcodes bodies by changing the associated social semiotics.

Experimental writing as assemblage
De Landa (2015) explains that Deleuze and Guattari give half a dozen different definitions of the concept 'assemblage' (p. 1). My work in this paper is guided by De Landa's efforts (2006,2015) to bring together the differences under a single workable concept. The aspects of written assemblage I want to emphasise are internal organisation and co-functioning. As we saw above, the double articulation comprises the form and substance of content and form and substance of expression. De Landa (2006) adds to this, 'A second dimension characterises processes in which these components are involved: processes which stabilise the identity of the assemblage (territorialization and deterritorialization)' (p. 19). De Landa adds to this a third dimension, relating to an assemblage's linguistic resources, or ability to capture the material world through coding (p. 19).
These ideas propose a very promising line of inquiry for thinking about experimental writing. I will sum up the main ideas to be carried forward to the next sections: Firstly, writing is not a transmissive practice but asserts its own identity as text and does its own work as textual assemblage. That is, a text distributes bodies (as with Foucault) but also is a body. Secondly, writing asserts both its own cohesiveness as textual territory, and its own liability to slip, divide, become other (territoralisation/ de-territorialisation). Thirdly, writing as assemblage effaces what we think we know and re-captures social reality as code.

Experimental text as assemblage: Overview and explanation
Though Deleuze and Guattari caution against the proliferation of models (1987, p. 499), I am mindful of the need to make the experiment understandable as well as affective. I therefore give an overview of the different sections of my text and explain the particular effects I will be aiming to achieve in each section.
(1) The self-assertion of the text. Here, I will approaching the text as a creative space. This will be in response to a snippet of data from a small research event, outlined below. The data are treated as an 'attractor' in De Landa's (2006) sense, mirroring its use in classical physics; i.e. not as a representation. I will be 'thinking aloud' to orient myself to the possibilities of text, in St. Pierre's (2016) sense of trying to think towards Deleuze and Guattari's flattened ontology through the medium of words. I will be trying to articulate the creative and radical potential of thinking in this way.
(2) The work performed by text. Here, I will be approaching the text as a generative space, in the sense that an engine generates movement. I will be theorising how words and signs give access to the data, prompted by discussion of Deleuze and Guattari's concept, 'regimes of signs' .
(3) The text as territory. Here I approach the text as integrated space. I will be affirming the internal coherence and cogency of the text by consolidating ideas already introduced. Bullet points are used to emphasise internal coherence. There is a line of argumentation intended to anticipate possible objections. (4) The text as dissolving territory. Here, I approach the text as dissolving space. I wonder aloud about how best to classify my text. Here, the text begins to concede its theoretical territory and I consider other possibilities.

Why is this a philosophical experiment?
The concepts 'diagrams' and 'diagrammatic' are very significant in Deleuze and Guattari's writing about signs (e.g. 1987, pp. 141-148). These concepts are not easy to grasp through conventional philosophical arguments. I will use my experimental text to come at them from a different angle. Despite its experimental appearance, this approach is distinctly philosophical in Deleuze and Guattari's sense of that word, because it addresses a field of problems defined by a concept (Deleuze & Guattari, 1994, p. 16). In this paper, my field of problems is defined by the concept 'diagrams' . A concept is defined as possessing 'variations according to their neighbourhood ' (1994, p. 20), while the purpose of philosophy is to 'map virtual potential as suggestively and productively as possible' (Holland, 2013, p. 20). The concept 'diagrams' will be mapped as imaginatively as possible, not logically deduced.
Thus, my purposes are, firstly, to map the concept 'diagrams' as it appears in a neighbourhood of problems generated by a data snippet from a small conference, and to do this by harnessing the internal dynamics of written text. Secondly, to generate an experimental text around the 'possibility space' (De Landa, 2006, p. 29) of the concept 'diagrams' , while attending to both the creative potential and writerly uncertainties arising from possibility spaces.

The small conference
The conference took place at a university in the north-west of England. It involved researchers from a number of international universities, school pupils, practising teachers, officials from various educational bodies and others. The purpose of this gathering was to debate and explore issues pertaining to global citizenship. At two points over three days, participants were invited but not obliged to respond to a number of stimuli such as a mobile voice recorder, a wall for public writing, a box for posting comments and a carousel of projected images. Participants were free to record, write, send or express any kind of message or response. This approach was informed by recent adaptations of the idea of 'entanglement' in different fields of research, as a means of theorising the fluid interfaces between 'persons, things, trajectories, sensations, discourses, and more' (Pink, 2015, p. 41. See also Ingold, 2008). I was also interested in recent innovative practice that aims to trouble simplistic divisions between the subject and objects of knowledge (Adams, Kueh, Newman-Storen, and Ryan (2015)), particularly as conceived by Mulcahy and Morrison (2017), where space is perceived as a relational and affective, as well as physical entity. The idea was to create a relaxed, responsive and receptive environment of practising, as well as theorising, global citizenship. It was hoped to create a space in which individuals might be 'affected' by one another (Zembylas, 2009).

Self-assertion of the text (text as creative space)
A diagram was drawn by a participant at the colloquium on a slip of paper and anonymously posted in a box (Figure 1). Throughout this paper, I have been challenging the notion of using data as a proxy for the human, but I seem to be doing it here. De Landa's theorisation of assemblages (2006,2015), helps to explain this inconsistency. My diagram is not supposed to be referring to an externally constituted world, at least not exclusively so. My diagram is deployed as an 'attractor' (2006, p. 29), defined as an invariant within a dynamic space of change, a 'property's capacity to be unaffected by a transformation' (2015, p. 113). Here, I am moving away from external reference and thinking about persistent thresholds of influence operating in my text. I was troubled when I first read the data diagram. It did not tally with the received wisdom that adults possess more social power than children. I am searching for the human presence behind the words but there are other possibilities. The data diagram can also be read as a creative engagement with anonymity, a gentle subversion of this researcher's efforts to renounce symbolic violence. Reading it as a space of signs, rather than as a space of reference, greatly increases its generative potential. Here, we are concerned with the internal dynamics of code, with the signifier 'school student' punching through the signifying order.
I often use diagrams in my teaching. I have re-created the data diagram to illustrate the kind of format I might use to schematise relationships between objects (Figure 2). This tidied up version is typical of diagrams found in textbooks and manuals. I follow de Freitas (2012) in suggesting that Figure 2 imposes a flat metric onto the original that presupposes a cognitive construction of its meaning. This deadens it impact as material artefact, reducing its potential to disturb and disrupt. Figure 1 in contrast is pushing its way out into three dimensions. The scratchy handwriting gestures towards its human author. The clipped outline shows it is a captured image. The shadows evoke a physical light source, revealing it as an object in space. It can be read back and forward at different speeds and intensities, so that different questions emerge about the act of reading (How should I look at it? What image of thought do I need?) I therefore tentatively propose a reading of Figure 1 as diagrammatic in Deleuze and Guattari's sense (1987). This is explained by De Landa (2015) as defining 'the structure of a possibility space' (p. 122). One way of understanding 'possibility space' , De Landa says, is to think of an 'ideally continuous cosmic plane' and think of it as broken up into actual, historical assemblages (pp. 108-109). These transformations are mapped through collections of lines, some rigid (with a high degree of territorialisation), some supple (low degree of territorialisation), while still others act as lines of flight, marking the directions along which an assemblage can become deterritorialised. (p. 109) The map of these lines is the attractor's diagram.  This idea relates to Deleuze and Guattari's concept, 'regimes of signs' or 'semiotic systems' (1987). These are the systems of signs that circulate around language, augment its functioning and are its very 'condition of possibility ' (1987, p. 140). For example, in a conversation a speaker uses body language, facial expressions and hand gestures in addition to words, vocal variation and so on. These semiotic apparatus might reinforce the speaker's verbal message (as classified under the signifying regime) or otherwise act to de-centre, confuse or complicate it (classified under post-signifying and counter-signifying regimes), depending on the interlocutor's reaction.

The work performed by the text (text as generative space).
The internal dynamic of a mixed semiotic can be regarded as generative when a form of expression is available that can accommodate the interplay of the different regimes (Deleuze & Guattari, 1987, p. 145). The data diagram represented in Figure 1 can be regarded as generative in this respect. It enables speculation about how the empirical situation was coded and how participants might have worked with, through or against this capture.
In theory, this series of transformations between different semiotic systems could be traced as a diagram (1987, p. 146). The mixed semiotic can be regarded as diagrammatic when 'we can conceive of this immanent plane in our minds by mentally forcing all movements of deterritorialisation to their extreme threshold' (De Landa, 2015, p. 109). At this limit point of thinking, the duality of signs (expression/content) begins to collapse (Deleuze & Guattari,, but other possibilities arise. We can begin to think of the data diagram as an assemblage, rather than as an image of something. Its incidental features (shadow, blurriness, scratchy writing) are thereby regarded as carrying ontological weight equal to its other components (message, meaning) and are equally implied in its co-functioning as space of possibility. Thus, its impact as a possibility space is greatly enhanced.

The text as territory (text as integrated space).
So far, the text has addressed the following important problematics about how we can deploy innovative writing in philosophical education research.
(1) The relationship between empirical work and the immanence of text.
(3) The difference between flattened diagram space and possibility space.
(4) Relationships between text understood as mixed semiotic and extreme abstraction.
In addition to deploying these creative and generative elements, I will consider the signifying semiotic in more detail in relation to education research. I said I wanted to answer the 'so what' question by generating a schema for other researchers to draw upon. This demands signification and representation, at least to some degree. Deleuze and Guattari (1987) point to the despotic origins of regimes of signification (p. 115) with its modality of infinite debt (p. 113). We can never repay the debt we owe to the signification despot-god; our day-to-day perceptions are structured for the pursuit of a final meaning or resolution that never actually appears (p. 112).
It could be argued that I have also avoided coming to firm conclusions in this paper. Nevertheless, I would argue that my paper is not structured in such a way as to lead my reader to expect ultimate resolution or significance. Elsewhere, Deleuze and Guattari argue, We should not forget that the strata rigidify and are organized on the plane of consistency, and that the plane of consistency is at work and is constructed in the strata, in both cases piece by piece, blow by blow, operation by operation. (1987, p. 337) The key point here is that signification is not an ultimate state or end destination, but a direction of travel at varying speeds and intensities. A regime of signs is subject to deterritorialisation pulling in both directions-'down' from abstraction towards lived particularity, and 'up' from the semiotic transformation towards the absolute threshold of abstract thought. To sum up, and returning to the methodological problematics identified at the start of this paper, signification can be theorised as an approach as opposed to a total effect, and as only one of several semiotic effects at work in a piece of writing.

The text as dissolving territory. (Text as dissolving space).
What is the potential for thinking the possibility space as pedagogy and as educational philosophy? My writing experiment is directed at both these fields. My purpose has been to identify a conceptual vocabulary of textual assemblage and exemplify how it can be used.
In What is Philosophy? The fields of philosophy and pedagogy are connected in the notion of concepts. Deleuze and Guattari write, 'philosophy is the discipline that involves creating concepts' (p. 5), while pedagogy is what safeguards philosophy from debasement through commercialisation (p. 12). This implies a view that learning, and by extension, teaching, cannot be reduced to the 'reproduction of linguistic capital' in the 'service of state standards' (de Freitas, 2012, p. 598). Deleuze and Guattari write, Concepts are not waiting for us ready-made, like heavenly bodies. There is no heaven for concepts. They must be invented, fabricated, or rather created and would be nothing without their creator's signature. (1994, p. 5) For learning to take place, the learner must fabricate thinking in their own way to address a specific field of problems. The role of the educator is to harness this creative potential in the pedagogical dynamic.
… we learn nothing from those who say: 'Do as I do' . Our only teachers are those who tell us to 'do with me' , and are able to emit signs to be developed in hetereogeneity rather than propose gestures for us to reproduce' . (Deleuze, 1994, p. 26) This paper's textual experimentation is both pedagogical and philosophical, in a particular way. It is addressed to a pedagogised 'other' presumed to be reading the text. It shows the addressee a way of emitting signs differently, through its own articulation of linguistic difference. It invites the reader to forge their own lines, thresholds and intensities in language. Yet uncertainties still remain (have I experimented hard enough? Is the result recognisably of pedagogic use?). My paper presents a schematic approach to linguistic assemblage, but I believe there is scope for much more research/experimentation in this area. One possible approach would be to look much more closely at linguistic texture, for example, the effects created by my deployment of active and passive voices in the text (e.g. 'I said…' 'It could be argued'). These variations work to position and assert my statements in particular ways. Developments in this direction would invite a move towards Deleuze's work The Logic of Sense, particularly its central concept, 'sense' . This concept disappeared from Deleuze's later work, but related concepts such as 'event' survived, and many of the supporting ideas could be fruitfully applied to the kind of experimentation I am advocating here (Lecercle, 2002).

Brief evaluation
Every assemblage is unique, but I think it is worth attempting to evaluate the experiment, partly to flesh out the suggestion made at the end of the last section (but not introducing new concepts from The Logic of Sense). I did not attempt to generate an argument in the conventional sense of ever-weightier deductions bound together by relations of logical entailment. I do not see this as either a strength or a weakness; it was enforced by the extreme abstraction of the subject matter ('diagrams' and the 'diagrammatic'). I do think that writing towards the limits of signification brings certain freedoms. For example, there is increased dependency on images and analogies, each of which tips the discussion towards its own zone of influence. For example, De Landa's notion of 'attractors' comes from the field of classical physics and pulls the writer in that direction. It is also possible to substitute images whose relationship to the idea is something other than through direct reference. So, in attempting to explain the 'diagrammatic' , I was able to move between an image of a diagram in a photograph, its approximate double, a mathematical theorisation of diagrams, and using 'diagrammatic' as a reference point for abstract thought. Each of these contributes something to the abstract idea and I believe heightens the sense of working within an assemblage of co-ordinated and co-functioning parts.
As alluded to in the section immediately above, certain techniques position and affect the arguments and might be worthy of much fuller analysis. I think it is important to be able to express doubt and hesitancy in ones advocacy of a point of view, as I do in sections one and four. One of the effects I find interesting is associated with the data (or 'attractor') in the text, which inhabits a liminal space, somewhat within and somewhat outside my frame of reference. Grappling with the presence of data is a particular challenge and a microcosm of the problematic nature of empirical reality in a Deleuze/ Guattarian frame. This form of writing allowed me to defer final judgment on its ontological status-this judgment would have been harder to avoid in a more conventional philosophical text requiring the definition of all terms in advance.
I also think it is important to be able to express certainty whilst acknowledging that one is only temporarily or provisionally certain. In the section 'text as territory' , I enjoyed writing the bullet points. The physical pressure of my finger on the key was pleasurable! Most academic texts deploy a rhetoric of certainty, but I do not think this is in any way misleading or unjustified. It is important to recognise that certainty, like the assemblage itself, is a direction of travel that will inevitably lead the writer elsewhere. Finally, I think the overall effect is cohesive. For example, I did not have to 'over-commit' to a particular angle on 'regimes of signs' or 'diagrams' . I was able to touch on these when it seemed suggestive or productive to do so. This seems very helpful when writing with Deleuze and Guattari as the ideas are already difficult and dispersed across many different passages, images and formulations.

Conclusion
We need to keep experimenting with writing to meet the creative possibilities of Deleuze and Guattari's flattened ontology. This paper presents one such experimentation. The experiment is explained, assessed and theorised in a non-reductive way, making new ideas and creative possibilities available to researchers in the fields of education and philosophy.
The paper meets and accepts the challenge recently articulated by Elizabeth Adams St Pierre of 'thinking with' Deleuze and Guattari. I have argued and demonstrated that this can be done by thinking with and through writing text. The paper developed a theorisation of 'assemblage' specifically adapted to a text's internal dynamic and co-functioning. The key ideas were that a text asserts and maintains its own identity while performing its function. These ideas were demonstrated in the experimental text. The paper also provided a summary of what the experiment was intended to achieve and an evaluative assessment of its success. Thus, in addition to its innovative qualities, the paper also provides a rich, conceptual language to support other researchers who might want to experiment along the same lines.
Finally, the paper indicates a future direction for researchers interested in the generation of linguistic effects. Future researchers might want to use Deleuze and Guattari's 'regimes of signs' , or Deleuze's The Logic of Sense, to source ideas for pushing these innovations.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes on contributor
Christopher Hanley is senior lecturer in education at Manchester Metropolitan University. His research interests include the philosophy of education, pedagogy and curriculum.