Introducing the ‘conceptual archive’: A genealogy of counterterrorism in 1970s Britain

Abstract This article contributes to an ‘historical turn’ in security scholarship. It addresses imbalance in security studies’ attention to historical empirics, and argues against notions of temporal disjunct prevalent within the discipline. I employ a genealogical framework to clarify the interpellation of past and present; and I introduce the ‘conceptual archive’ as a lens for pursuing that interpellation in research. My thesis on the ‘conceptual archive’ represents a twofold contribution. Firstly, a conceptual contribution: I advance the ‘conceptual archive’ as a way of thinking about past-present interpellation (specifically, existing conceptual logics’ remodelling in arguments justifying new practice). Secondly, an analytical contribution: I propose the ‘conceptual archive’ as a tool for doing genealogy (a research programme with historicising promise, but one suffering nebulous operationalisation at present). I use the field of terrorism studies as an entry-point to these contributions: adopting a mixed-methods research design to trace British counterterrorism practices’ roots within an ‘archive’ of logics on Northern Ireland. I find 1970s British governments remodelled long-standing ‘archival’ vocabularies in their arguments for new security provisions: framing exceptional practices according to an accepted fabric of concepts.


Introduction
Recent scholarship has called for greater historicity in security studies. Critics have noted the 'fraught' , 1 'amnesiac' , 2 and even 'intimately adverse' 3 relationship between security research and the historical past: thanks to which a 'broadly non-historical centre of gravity' has 'long prevailed in the field' . 4 Writing in this journal, for instance, Faye Donnelly and Brent Steele found it 'surprising' that 'very few' studies had 'explicitly foregrounded the intricate interrelations that exist between history and security' 5 -an observation affirmed by journal editors in their invitation to I proceed to demonstrate this framework's empirical substance, adopting a mixed corpus linguistic/discourse analytic method to explore a particular evolution in security -'new' counterterrorism practices in 1970s Britain. I establish the continuity and parameters of an 'archive' of logics in British political discourse (concerning the 'problem' of Northern Ireland, and prevailing since the latter's creation in 1920). I then identify this archive's footprint in 1970s ministerial rhetoric justifying 'unpalatable' departures in domestic security. I end with reflections on these findings' import. I suggest the conceptual archive realises efforts to 'broaden the historical imagination' 11 in security studies: contributing a platform for thinking historically, and an analytic for doing historical research.

Security studies' dual problem with the past
Existing security literatures ('orthodox' and 'critical' alike) 12 suffer a twofold ahistoricism -the first, and most straightforward, feature of which is security studies' temporal myopia. This is a question of temporal mandate: a 'presentism' , 13 'recentism ' , 14 or 'tempocentrism' 15 in which findings and theory are drawn from narrow (temporal) horizons. That is, a horizon of cases from the recent past, or courte durée -relegating the deeper past to a 'peripheral' 16 zone within the discipline.
Past reviews find security scholarship suffers an 'extraordinarily "lopsided" focus' 17 on the recent past: with studies 'frequently [operating] an unhelpfully abbreviated 21st-century lens'; 18 or, less generously, exhibiting 'an almost complete lack of historical awareness' . 19 Beyond manifesting a research 'gap' , this temporal myopia constrains security studies' conceptual scope. Analytical frameworks drawn from present empirics risk misinterpreting past phenomena: misconstruing contemporary dynamics as 'universal historical occurrence[s]' , 20 and projecting these across time without sensitivity to context (paralleling the 'Eurocentrism' whose critique is well established in security studies 21 -though in time, rather than space). More importantly, they may also misread phenomena from the present -whenever these phenomena are not captured in narrowly contemporary data. As, for instance, in the phenomenon this article theorises: past conceptual logics' footprint in discourses legitimising new practice. This dynamic, so central to present possibility, would be impossible to ascertain from tempocentric empirics alone -since these couldn't tell us which past logics a present discourse speaks to. To continue the 'Eurocentrism' analogy, this is the equivalent of blindness to 'whiteness' in research analysing European phenomena in isolation from their 'imperial conditions of possibility' 22 -which tells us as little about the European, as the non-European experience.
The second, and more significant, feature of security studies' dual problem with the past emerges from scholars' response to the first. Some scholars rightly reject security studies' temporal myopia: suggesting 'analysis … has been compromised by the failure to adequately historicise and contextualise [concepts]' . 23 Instead of projecting 'current concepts onto historical practices' , 24 as in tempocentrism, these scholars emphasise the 'contingency' of different periods: 'the contingent and mutable constitution of social forms' , 25 and the 'distinctiveness of social phenomena [across] 12 Harmonie Toros, "'9/11 is alive and well" or how critical terrorism studies has sustained the 9/11 narrative' , Critical Studies on Terrorism, 10:2 (2017), pp. 203-19 (p. 211). 13  history' . 26 Such emphasis invites assessing mercenary violence in quattrocento Italy or contemporary Libya, for instance, 27 on their own terms: in 'the specificity of … their temporal and spatial contexts' , 28 instead of 'assuming any conceptual continuities over time ' . 29 This approach has its strengths. It resolves tempocentric theorisation: refuting past empirics' misinterpretation through presentist frameworks, by 'accept[ing] that many aspects of the past are significantly different from those of our own time' . 30 But there is also 'a danger of over-correcting' 31 here. Contingent approaches' concern with 'context and indeterminacy' 32 risks throwing the historical baby out with its bathwater: engendering 'a sovereign politics of time' by inscribing 'temporal borders' 33 between periods. In stressing past times' 'radical "otherness"' , 34 contingent approaches institute artificial barriers between them. Scholarship operationalising the past in terms of 'discontinuity, contingency, and particularity' 35 ends up 'cutting' 36 history into a series of disconnected chunks: a 'butterfly of contingent hiccoughs' , 37 rather than an intricate and interpellated whole. This erection of temporal sovereignties manifests a reverse presentism: siloising historical times as fundamentally different to, and lacking connection with, each other. In so doing, it overlooks 'broader processes, sequences and plots' , 38 which make the past meaningful -internally, and to the present. Past time is not divisible. On the contrary, it 'flows, events overlap and break against each other like waves on a shore' . 39 Attempts to 'freeze' 40 history through notions of contingency marginalise temporal intersectionsand diminish the past's footprint within the present.

A solution? Genealogy
Such is security studies' dual problem with the past: either an 'obsession' 41 with the present in empirical analysis, constraining the discipline's conceptual scope; or, a 'fetish' for 'the particular and the exceptional' 42 negating past-present interpellation. All is not lost, however. For 'there is another literature' containing 'the potential to push these limits' . 43 This is the literature of security studies' 'genealogical turn': 44 including important works redirecting the scholarly gaze into a deeper past, without invoking temporal disjuncture. 26 Cello, 'Taking history seriously in IR' , p. 238. 27 Malte Riemann, "' As old as war itself "? Historicising the universal mercenary' , Journal of Global Security Studies Genealogy's relationship with security studies begins with the discipline's 1990s 'linguistic' and 'practice' turns. 'Closely linked' 45 to each other, these turns frame discourse/language as a 'field of social and political practice' , 46 which brings objects of security into being, thus shaping political relations. According to linguistic/practice theorists, political phenomena like 'security' are not 'given and self-evident' 47 -'pre-formed and ontologically separate' 48 from discourses by which they are imagined/evoked/interpreted. On the contrary, they are 'essentially contested' . They contain 'multiple meanings' 49 -which 'no amount of discussion can possibly dispel' , 50 and whose salience depends on 'perception' 51 or 'ideological [and] moral element[s]' . 52 Given security phenomena's essential contestability, linguistic/practice theorists suggest their dynamics have less to do with 'reality' 53 than the 'deliberately and meticulously composed … myths and forms of knowledge' 54 that constitute them. Security is 'literally inconceivable' 55 beyond language. As such, speaking 'security' involves an 'almost magical power' 56 to 'shape the terrain upon which [political] struggles take place' 57 -with security's discursive imagination having material effects for political relations (as in securitisation theory, which proposes the 'practice' of 'saying "security"' moves conflicts 'above politics' , 58 as a precursor to their closure). Security, in this sense, is 'as much about ideas as it is about guns' . 59 'What happens in someone's mind, or in the minds of a series of individuals' 60 is not ornamental to material realities. Instead, concepts 'have teeth' . 61 Discursive practice is 'the political action par excellence' , 62 forming 'a necessary part of our study' . 63 Genealogy's value to that study is straightforward. According to genealogical traditions, the concepts constituting discursive practice do not emerge in isolation from each other. Rather, they exist in a conceptual ecology. Per Friedrich Nietzsche, concepts are not something arbitrary, something growing up autonomously, but on the contrary grow up connected and related to one another … however suddenly and arbitrarily they 45 Jérémie Cornut, 'The practice turn in International Relations theory' , Oxford Research Encyclopedia of International Studies appear to emerge in the history of thought, they nonetheless belong just as much to a system as do the members of the fauna of a continent. 64 Exploring politics' discursive constitution requires locating concepts within this ecology. Discourses don't emerge overnight. They evolve through 'descent' 65 -acquiring shape and traction from the existing ecosystem. As, for instance, in Nietzsche's thesis on the origins of 'guilt' in the 'material concept of "debt"': 66 the former being a reinvention of the latter. This 'palimpsestuous' 67 take on discourse directs the scholarly gaze into the past: tracing contemporary discourses' descent from prior ecologies in a 'history of systems of thought' . 68 An emerging ('but still very small') 69 community of security scholars have undertaken this task. These scholars note how contemporary discourses/practices around counterterrorism, 70 confinement, 71 or peacekeeping 72 remodel 'repertoire[s] of cultural idioms, categories, and narratives' 73 from the past. They establish ways 'our current [security] order has been shaped by, or continues to be shaped by' 74 'historical matrices' 75 of discourse -suggesting past conceptual ecologies remain 'overlaid … readable, or operational' 76 in contemporary practice (including especially ecologies associated with colonialism 77 and gender). 78 Such scholarship affords a path beyond security studies' dual ahistoricism: inviting attention to past empirics, while retaining sensitivity to temporal interpellation (past ecologies' 'deep, sticky' 79 footprint in contemporary practice).
However, there remain problems in pursuing this path. Foremost among these is genealogy's nebulous application. Genealogy is an 'unsystematic' 80 tradition: with advocates having undertaken 'comparatively little systematic work or meta-theoretical reflection on genealogy as a method of political theory' , 81 and with 'the question of how we should go about investigating and interpreting the past [being] rarely asked, let alone answered' . 82 Genealogy's 'really frustrating' 83 lack of methodological systematicity is by design. Genealogists have avoided 'produc[ing] plans for some future building' , 84 preferring instead 'a playful, improvised and disruptive mode which consciously avoids offering a simple and coherent account' . 85 This flexibility has merit, insofar as it prevents 'warping empirical materials by subjecting them to a framework whose contours were developed elsewhere' . 86 Nevertheless, the instinct for analytical freedom has a cost. It 'puts [genealogy] at a disadvantage within the research community' 87 by constraining possibilities for its application. Genealogy's opaque/nebulous operationalisation inhibits widespread use; 'with such vague guidelines … it is little wonder that political scientists have, with very few exceptions, not taken up the challenge' . 88 Realising the past-present interpellation genealogy theorises requires some framework clarifying its practice -illuminating 'the dark tunnel of genealogical method' to 'make it "usable"' . 89 This is where my 'conceptual archive' comes in.

The conceptual archive
The conceptual archive is a way of thinking about past-present interpellation, and an analytic for pursuing that interpellation in practice. It channels genealogy's core theoretical premises: locating discursive practices within their conceptual ecology, with present discursive possibilities being shaped, animated, and sustained by past conceptual inheritances. And it systematises that theory of discourse across two levels: 1. A search for 'archival' logics pertaining to a discursive field; and 2. An assessment of how arguments for new practice remodel those logics.
The first part of this framework begins with a simple intuition: that basic sense-making rules are a prerequisite for discursive exchange. As Antje Wiener suggests, 'understanding is never unmediated' . 90 On the contrary, communicative agents' adherence to a common frame of reference is a precondition for 'mutual intelligibility' . 91 Per Michel Foucault, 'if language exists' it is only because 'below the level of identities and differences there is [a] foundation provided by continuities, resemblances, repetitions, and natural criss-crossings' . 92 This foundation makes sensible discursive exchange possible, providing a 'central authority structure' to 'halt the fluidity of terms and make language meaningful' . 93 Without it, discursive exchange would descend into a 'chaoscosmos' 94 of incoherent/disconnected fragments. Conversely, the necessity for a common frame of reference ensures a degree of continuity in discursive practice. If agents consistently return to collective frames to assure intelligibility, then discursive structures will remain stable over time: setting parameters for what agents can sensibly say, across generations of practice (a system of 84  Rules for present discursive practice are inherited from the past. Hence, the 'archive': an ensemble of 'historical a priori' 96 for discourse, or reservoir of inherited logics -into which agents dip when making interventions, and against which audiences refer to interpret them. This archive is the equivalent of a familiar tune, engrained knowledge of which 'is exercised in humming or playing it'; 'in recognising and following the tune, when heard'; and 'in noticing errors in its misperformance' . 97 The first step in a 'conceptual archive' analysis is to establish this tune's parameters, looking for familiar logics' consistent recirculation/reappearance (what I refer to as keyword/collocate 'undulation' or 'porpoising' in my empirical analysis). The second step is to identify these parameters' footprint in arguments justifying new practice -establishing 'resonance' between legitimising discourses and the archival 'tune' .
This second stage derives from a further intuition: that all evolutions in practice require some argument by which to legitimise them. Per Wiener, again, unfamiliar forms are 'contested by default' . 98 Those forms that can't be legitimised in the face of such contestation 'cannot be pursued over the long haul' . 99 In short, 'legitimation is … an imperative, not a mere nicety' 100 for new practice. But the acquisition of legitimacy is a relational process. Legitimacy is not dictated by the proposer, but negotiated with their audience. It is not the case that 'anything goes' 101 in legitimising practice. Rather, prospective legitimisers must work with stakeholder expectations: cutting their arguments for new practice from an established conceptual fabric. Successful legitimising discourses are those 'resonating' with that conceptual fabric. Namely, employing rhetorical 'frames' ('predefined structures [that] already belong … to the receiver's knowledge of the world') 102 conforming to the 'existing range of favourable evaluative-descriptive terms' 103 within a conceptual ecology. Unsuccessful discourses are those that depart radically from that ecology: lacking rhetorical purchase, 104 or even being 'beyond comprehension and reason' . 105 The archive is the conceptual ecology in which discursive agents operate. Agents who successfully assimilate archival logics enhance possibilities for new practice. Equally, agents speaking in abstraction from archival logics will 'meet with a rude reception' , 106 and are 'likely to fail to achieve [their] goals' . 107 As Quentin Skinner put it, these constraints oblige 'every revolutionary … to march backward into battle': 108 modelling past logics to guarantee sense and traction, even in communicating evolutions of practice. The archive operates as a scope condition for evolutions in practice. Since 'any course of action is inhibited from occurring if it cannot be legitimated' , 'any principle 95  which helps to legitimate a course of action must also be amongst the enabling conditions of its occurrence' . 109 Having established the existing archive's logics, then, my framework further suggests identifying their 'footprint' in discourses enabling change -noting how agents remodel those logics in legitimising new forms. I describe this footprint in terms of 'resonance': synergies between archival logics, and those of a legitimising discourse. This approach to discursive practice resembles 'facilitating conditions' in securitisation theory -the 'conditions under which the [securitising] speech act works, in contrast to cases in which the act misfires or is abused' . 110 Where the archive innovates on 'facilitating conditions' is in the specific relationship it establishes between past and present in imagining them. That is, the multigenerational archive of sense-making logics: handed down from the past, and constituting the 'conditions' to which discursive agents must speak. This temporal interpellation unites past and present in an assessment of evolving practice, with archival ecologies being a condition of present political possibility.
This framework also unites change and continuity (resolving discomfort 111 with structuralist models depriving discursive practice of agency). On the one hand, the conceptual archive is concerned with continuities, per its intention to connect past and present. Instead of assuming agents speak freely, in abstraction from their conceptual ecology, the archive looks to 'parameters of sensemaking' 112 to which agents conform for intelligibility/purchase. As in practice literatures on 'habit' , such conformity 'anchors actors' perceptions, attitudes, and practices toward the status quo' . 113 The archive reproduces itself continuously, as a condition for meaningful discourse. On the other hand, the archive does not eliminate possibilities for change or agency. Rather, it anticipates them: recognising individual efforts to introduce new practice, but connecting these to 'vocabular[ies] already normative within [their] society' . 114 The archive does not deny authorship. Instead, it proposes sensitivity to ways authorship takes shape; with agents adding value to their designs by 'appropriating' the archive's 'rich resources' , 115 as much as being limited by its constraints.
The archive embeds this duality in its twofold analytical framework. First, exploring conditions for discursive exchange by establishing archival parameters. Second, clarifying agents' operations within those parameters (remodelling archival logics in legitimising new practice). In Pierre Bourdieu's words, this is a question of 'invention within limits' . 116 The conceptual archive assumes agents communicate their own ideas, just not in a conceptual fabric of their own invention -per the classic Marxist bridge 'man makes his own history, but he does not make it out of the whole cloth, he does not make it out of conditions chosen by himself, but out of conditions found, given, and transmitted from the past' . 117

Britain's 1970s counterterrorism moment
This article introduces the conceptual archive as a way of thinking about past-present interpellation, and a tool for pursuing that interpellation in research. It does so by considering a particular case of security practices' evolution: 'new' counterterrorism's emergence in 1970s Britain. I chose to consider counterterrorism because, of all security literatures, terrorism studies suffers the greatest ahistoricism. 118 And I chose the British case as an entry-point to my contribution. Not only, because the UK was the first state to transition to an explicit 'counterterrorism' paradigm 119 (in whose practice it subsequently 'stands apart' , 120 or claims 'profound expertise'). 121 But, also, because of heightened pressures to justify evolving security provisions in Britain's common law traditionthrowing conceptual archive dynamics into sharp relief.
Counterterrorism 'began' 122 for Britain in the 1970s, with the Northern Ireland (Emergency Provisions) Act 1973 and Prevention of Terrorism (Temporary Provisions) Act 1974. The Emergency Provisions Act (EPA) introduced Northern Ireland-specific measures, including: nonjury courts for scheduled offences; non-attendance for witnesses/defendants at trial; detention without charge; proscription of political parties; reductions in the onus of proof for prosecution; stop-and-search; bail restrictions; and other security provisions. The Prevention of Terrorism Act (PTA) transferred detention and proscription powers from Northern Ireland to the wider UK, supplementing these with additional powers of exclusion for terrorist suspects (an executive prerogative of the Home Secretary, requiring no evidential basis). 123 PTA also extended maximum detention periods from 48 hours to seven days. Although individual EPA/PTA powers were not unprecedented in British history, 124 their assembly in the specific name of 'terrorism' was. Before 29 November 1974, the UK Parliament had never passed a law featuring 'terrorism' in its title. Since then, it has passed 260 -one piece of terrorism legislation every two months. 125 Such was the novelty of mid-1970s British security. But how might Britain's 1970s counterterrorism moment illuminate my thesis on the conceptual archive? The British case's particular value lies in its common law traditions, and their associated rhetorical pressures. Others have explored counterterrorism's proliferation in Roman law polities, through activation of constitutional 'state of exception' clauses -an administrative process involving less a 'politics of security' than its 'lasting eclipse' . 126 The UK, however, does not have a constitution. When the desire arose for disciplinary powers in the 1970s, therefore, British governments had to create new laws: and present these to critical stakeholders. Most obviously, critical stakeholders in Parliament -a necessary 'site' 127 for discursive practice in a common law context, whose support for new measures governments must win through carefully curated argument.
These rhetorical pressures were particularly sharp in 1970s Britain. Government ministers recognised the significant shift in security norms EPA/PTA provisions manifested. Ted Heath described EPA, for example, as a 'marked change in our practice' 128 -one Home Secretary Robert Carr thought would be 'highly controversial [and] extremely unpalatable' , 129 132 or 'lead to extreme difficulties in Parliament' . 133 Hence, Government developed a deliberate rhetorical programme by which to 'stage manage' 134 Parliament's initiation to counterterrorism. Central to this programme was a repurposing of conceptual logics already familiar to parliamentarians. Namely, logics pertaining to parliamentary discourse on Northern Ireland: my 'conceptual archive' . 1970s Britain represents a useful case to tease out the past-present interpellation my conceptual archive theorises. 1970s British officials faced pressures to justify new powers in Parliament -pressures throwing archival dynamics into illuminating relief. This 'spatial' case selection is simultaneously a 'temporal' one, however. Given my arguments against temporal myopia, why begin this study in 1920 (a longer timeline than most research, but not yet the longue durée)? In refuting common security chronologies, Harmonie Toros calls for empirically grounded research design, suggesting scholars use whichever timeline is 'most relevant to the subject at hand' . 135 In my case, 1920 is a relevant place to begin this article, as a reasonable origin for the specific 'conceptual archive' my study investigates; namely, the archive of British political discourse on Northern Ireland. Northern Ireland is 'only a recent invention' , 136 in both administrative and conceptual terms. The spatio-political entity 'Northern Ireland' did not exist until the Government of Ireland Act 1920. More significantly, the discursive concept 'Northern Ireland' only entered British discourse after 1920 -featuring merely three times in pre-1920 parliamentary debate (once each in 1861, 1904, and 1919), but 179 times in 1920. 137 The years between Northern Ireland's creation and British counterterrorism's emergence remain an appropriate horizon within which to generate insights on 'old' and 'new' discourses' intersection, therefore. In my case, the intersection between archival logics on Northern Ireland, and arguments justifying 1970s security.

Methodology
Before I elaborate these insights, a note on my mixed methodology. My analysis is based on a twofold assessment of data. Firstly, quantitative corpus linguistic assessment of material from the Hansard record of parliamentary debates. Secondly, qualitative discourse analytic assessment of UK Government documents from the National Archives. My 'corpus-assisted discourse analysis' builds on a growing trend towards 'methodological synergy' 138 in discourse studies: combining breadth of computational analysis 139 with depth of close reading. 140 This is not the only way to operationalise the conceptual archive; but it does demonstrate insights afforded by its two analytical levels.
I begin with corpus linguistic assessment of a new dataset of language on Northern Ireland from the period 1920-84 (from Northern Ireland's creation to the end of Britain's counterterrorism 'moment'). The 'Northern Ireland in Parliamentary Discourse' dataset (NIPD) is a 12,751,975-word corpus capturing all UK Parliament debates on Northern Ireland in the 65 years following its creation. I divide these 2,449 debates into 13 five-year blocks 141 (emulating similar historical discourse studies of prostitution 142 and race): 143 with a mean word count of 980,921, and with each block corresponding to a different 'chapter' in Northern Ireland's first decades (Northern Ireland's creation (1920-4); the Second World War (1940-4); the civil rights movement (1965-9); the Heath administration and Direct Rule (1970-4); etc.). This article advances a series of findings about language use over 65 years of parliamentary discourse; it does so by tracking change/continuity of specific measures between these blocks. The measures I use are Simple Maths Parameter (SMP) to calculate word keyness, and Mutual Information 2 (MI2) to calculate collocation.
'Keyness' is a tool for evaluating a corpus' 'aboutness' . That is, specific keywords' relative prevalence, as a proxy for corpus topicality. 144 SMP is one means of ascertaining keyness. It operates by comparing relative frequencies of all words in a 'focus' corpus, to relative frequencies of all words in a 'reference' corpus. 145 The formula's beauty is the ease of interpreting its findings. SMP quantifies keyness in 'simple' terms: if a word returns an SMP of 2.00, we can interpret it as being twice as prevalent in the focus corpus vis-à-vis the reference corpus. My keyness analysis in this article proceeds through two phases. First, I acquire a list of words that are externally key in NIPD: words that are substantially more prevalent in my 13 corpus blocks than in contemporaneous reference corpora of generic policy language. 146 I calculate external keyness for all words in NIPD: removing words with an SMP below 2.00, along with stopwords, to generate a shortlist of 324 external keywords. These 324 are words that were at least twice as prevalent in parliamentary language on Northern Ireland as in contemporary policy discourse (SMP>2.00). Capturing terms like 'problem' , 'anomaly' , and 'situation' , these keywords conceive Northern Ireland in a particular way. Namely, as an 'exceptional' and 'troublesome' space in the British political imaginary -an 'urgent' 'question' in need of 'solving' . I repeat this keyness procedure for my 13 blocks: comparing my 324 keywords' relative frequencies in each block to their relative frequencies in the whole corpus, to generate data on internal keyness. That is, how much more prevalent a keyword is in one period of NIPD, vis-à-vis NIPD generally. I use these calculations to create a diachronic keyness matrix: detailing change/continuity in my 324 keywords across 65 years of discourse on Northern Ireland. I employ this matrix as the basis for contentions on the conceptual archive (using standard deviation, an 'attractive technique' 147 for diachronic corpus linguistics, to measure keyness variation). If keywords exhibit significant stability across NIPD (SMPs 'undulating' around 1.00/low standard deviations) this would suggest the existence of a stable conceptual archive: a set of logics for thinking about Northern Ireland, consistent in parliamentary discourse across historical 'chapters' . 'Collocation' is a tool for ascertaining the 'meaning' of words in a corpus. Collocation metrics quantify meaning by calculating how often keywords appear alongside other words within the corpus -based on the Firthian principle that 'you shall know a word by the company it keeps' . 148 Collocation assumes word meanings emerge from contexts of their use. Frequent collocation of 'time' with words denoting value, for example, implies a semantic relationship 'time is money' . 149 Having ascertained change/continuity in my 324 keywords' prevalence across NIPD, I thus also advance an assessment of change/continuity in their meanings. I subsample relevant keywords from this group of 324 and track variation in their collocates over time. I use MI2 for this purpose. MI2 is a measure of exclusivity of collocation: comparing observed frequencies of words' co-occurrence with chance co-occurrence, if the corpus was reorganised randomly. A high MI2 implies a 'tight' 150 semantic relationship -one signifying the uniqueness of words' co-configuration. For reference, when applied to the BE06 corpus of modern British English, 151 the search parameters employed below returned MI2s of 9.18 for 'royal'-'family' and 9.26 for 'university'-'student' . 152 These are very strong collocations in British English. Scores above 9.00 can therefore be taken as evidence of very strong semantic relations in NIPD. I use collocation analysis to supplement findings on internal keyness. Whereas stable keyness scores suggest quantitative continuity in parliamentary language (keywords are used equally frequently over time), stable collocation scores imply qualitative continuity (they are used with the same meanings). Such qualitative continuity would support my claims about the existence of a conceptual archive in British political discourse.
Finally, I bring depth to corpus linguistic returns via qualitative reading of non-digitised UK Government documents from the National Archives. I explore how Government mobilised discursive logics from the conceptual archive in language justifying 'unpalatable' counterterrorism. In particular, I employ metaphor analysis to explore Government's reliance on an accepted vocabulary (the 'conceptual archive') in framing these provisions. Metaphors work by 'smuggling' 153 shared values into an argument: tying common vocabularies (the metaphor) into an argument's logic, and leaving uncomfortable details unelaborated. Metaphors enable a mechanical style of discourse, concealing contestable propositions within a mutually accepted conceptual fabric. I find Government's argument on counterterrorism employed multiple metaphors from the conceptual archive -portmanteau connectors between rhetorical propositions, which worked precisely by channelling a reservoir of logics for thinking about Northern Ireland.
Across both quantitative and qualitative analyses, I look for synergies between long-standing logics of parliamentary discourse on Northern Ireland (my conceptual archive) and Government discourse on counterterrorism (my evolution in security). The latter's constitution according to the former would substantiate my claims on the conceptual archive. Namely, that the conceptual archive operates as a scope condition for justifying new practice.
Genealogising 1970s counterterrorism I begin by tracking keyness returns for NIPD's 324 external keywords (words that were at least twice as prevalent in discourse on Northern Ireland as in generic policy discourse, 1920-84) across my 13 corpus blocks. If my hypothesis on the conceptual archive's stability holds, we should see limited variation in keyness returns over time (SMPs hovering around 1.00/low standard deviations)indicating these keywords are not especially prevalent in any single period of NIPD. But, rather, that  Across 324 keywords, mean internal keyness stands at 1.00 with a σ of 0.58. 55 per cent of keywords have a σ below 0.50, and only 5 per cent one above 1.50. Such minimal variation suggests the proliferation of a style for discussing Northern Ireland across its first seven decades … A notable feature of which was its peculiar reliance on temporally situated keywords (peculiar because these keywords must have been at least twice as prevalent in NIPD as in generic policy discourse to merit inclusion) -keywords such as 'history' (mean keyness 0.92/σ 0.32), 'moment' (1.05/0.32), 'emergency' (0. 82  Stable SMPs across corpus blocks (minimal variation in keyword prevalence/low standard deviations) are indicative of a 'conceptual archive' in British political discourse on Northern Ireland. Parliamentarians became habituated to a programme for speaking about Northern Ireland, over decades of debating its politics -conceiving these as 'problematic' and 'exceptional' . The boxplot in Figure 2, visualising SMPs for all 324 keywords across 13 corpus blocks, exemplifies this trend. Median SMPs (the central bar in each box) hover around 1.00 -with no significant upward/downward pressure in keyword prevalence between blocks. This is what I would expect the conceptual archive to look like: a steady recirculation of concepts in discourse, albeit with gentle keyness undulations over time.
Alongside its generic currency in parliamentary discourse, it appears the conceptual archive attains additional purchase at moments of heightened insecurity. Parliamentarians are particularly likely to invoke archival logics when facing challenging questions. Its familiar languages serve as a conceptual comfort blanket: furnishing preformed conceptual schemes by which to rationalise these questions (a 'repertoire or "tool kit" of habits, skills, and styles from which [to] construct "strategies of action"'). 163 Already, we see mutuality between conceptual archive and 'security' (parliamentarians employ familiar languages to make sense of insecurity, but it is precisely their sense of insecurity that directs them to those familiar languages). 164 I explore this mutuality in greater detail later. For now, let me return to corpus linguistic analysis.

The conceptual archive: Collocation analysis
promoted by Lord Chancellor Hailsham on EPA ('although these measures appeared draconian, there was a safeguard in that they … would remain in force for only 12 months'); 175 and Home Secretaries Roy Jenkins/Merlyn Rees on PTA ('I do not see the act as anything other than a temporary measure' 176 /'powers contained in this act are exceptional … a temporary infringement of civil liberties' .) 177 Temporary time resolved EPA/PTA's 'unpalatability' , by framing them as a temporary anomaly. Underlying liberal norms are unaffected by practices taking shape under an anomaly: the anomaly is provisional, but the underlying norms are transcendent. This logic appeared in briefing materials ('the bill makes no attempt to deal with the rules for normal times'), 178 speeches ('no-one will welcome provisions which are recognised to be unacceptable in normal times'), 179 and published reports ('recommendations to take effect only so long as the emergency … continues'). 180 Such temporary time arguments relied on a binary between 'exceptional' and 'normal' timesa binary in which the conceptual archive played a definitive role. In private correspondence, officials emphasised 'for [EPA] to command general support, we think it should be made clear that its provisions are intended for use only in an emergency situation' 181 (a comment Northern Ireland Secretary Willie Whitelaw annotated with 'agreed, Very important'). The familiar archival logic 'the Emergency' thus became a central part of Government's argument on counterterrorism: as the conceptual identifier distinguishing 'exceptional' from 'normal' times. For instance, in a speech by Whitelaw on EPA ('a temporary measure to meet an emergency situation'), 182 or Rees on PTA ('the act makes emergency provisions and is by its nature temporary, to cover the period of an emergency'). 183 This identifier attained rhetorical purchase thanks to its familiarity in parliamentary discourse. As noted earlier, 'the Emergency' was an established part of the conceptual archive -'becoming the norm' 184 in the ensemble of logics for thinking about Northern Ireland. When ministers invoked 'emergency' to substantiate their argument on 'temporary time' , therefore, they were invoking a concept with existing purchase in the parliamentary imagination -one from whom they expected limited demurral ('I think it unlikely any honourable member … would dispute the contention of the Government that an emergency exists in Northern Ireland at the present time'). 185 Temporary time was essential to 1970s' governments' argument for 'unpalatable' counterterrorism. It resolved contradictions of illiberal security practices in a polity priding itself on its liberal traditions, posing these practices as 'out of the norm with conventional legal provisions' . 186 This is the equivalent of Michelle Bentley's thesis on 'normative invalidation' -proposing discursive 'invalidation' as one means by which agents resolve 'normative dilemmas' associated with securitisation. 187 Temporary time represents a temporalised instance of normative invalidation: involving efforts to temporarily 'neutralise the relevance and/or meaning' 188 of Britain's liberal mythology, in the face of 'emergency' circumstances. And it worked precisely through resonance with the archive connected through recourse to 'present circumstances': a temporally situated euphemism, deriving its meaning and purchase from the archive. Ministers would not have been able to employ 'present circumstances' as a connector in their argument on counterterrorism, if these weren't a pre-established feature of parliamentary discourse -a portmanteau of assumptions about Northern Ireland's 'problematic' politics that could be stated as established fact: 'it is … a sad fact that normal legal processes are no longer adequate in the circumstances of Northern Ireland'; 198 'it is an unfortunate fact that … under these circumstances the normal agencies of law enforcement have been unable to cope ' . 199 As with temporary time, Government's argument on EPA/PTA worked because it resonated with conceptual archive logics. In this case, remodelling euphemistic 'present circumstances' to bypass challenging detail.

Conclusion
This article introduced the conceptual archive as a way of thinking about past-present interpellation, and a framework for pursuing that interpellation in practice. My argument on the conceptual archive highlights discourses' fundamental stickiness over time: with consistent conceptual logics being a prerequisite for sensible exchange, and with arguments for new forms acquiring purchase by remodelling those logics. The 'archive' is the conceptual ecology in which discursive agents operate. Innovating agents, who 'understand the benefits' accruing to 'those who can rework … underlying narratives' , 200 must speak to that ecology, even in advocating change -cutting arguments for new practice from a familiar conceptual fabric. This means evolutions of security will always carry the footprint of archival logics. Identifying that footprint can put past and present in communication: revealing how past conceptual inheritances shape, animate, and sustain present political possibilities.
Applying this framework to British counterterrorism's 1970s emergence, I found arguments for the new security paradigm worked by remodelling logics from an 'archive' of discourse on Northern Ireland. Parliamentary discourse on Northern Ireland between the latter's creation and Britain's 1970s counterterrorism moment centred on an ensemble of logics relating to the 'problem' of Northern Ireland: a 'habit of mind' 201 or 'frozen regime of thought' 202 structuring debates on Northern Irish politics across their first decades. This ensemble also enjoyed centrality in rhetoric justifying 'new' counterterrorism. 1970s ministers recycled 'frozen' archival logics when legitimising 'unpalatable' security provisions. Specifically, ministers recycled logics concerning Northern Ireland's temporal liminality: justifying exceptional derogations from liberal norms, by connecting these to Northern Ireland's permanent 'emergency' and peculiar 'circumstance' . Such findings substantiate recent insights on the 'will to time' 203 in political relations: 'recognis[ing] time as fundamental' 204 to new practices' evolution. Moreover, they illuminate my 'conceptual archive' framework's empirical scope: demonstrating intersections between a tradition of logics structuring political discourse, and arguments advanced by innovating agents to justify new departures.
stuff we felt sure we could pour into it' , 215 and advocate analyses combining textual with non-textual data. Establishing whether the conceptual archive operates multimodally could expand its scope. For example, exploring security's spatial manifestations, and how these also channel archival logics.
Finally, analysis in this article rests on material from 'elite' 216 sources: the UK Parliament/Government. This was a methodological choice, based on data availability. But it also holds implications for the article's attention to discourses' relational character -and oppositional agency's role in shaping conceptual archive logics. Per Foucault, there is 'no free, neutral, independent statement … a statement always belongs to a series or a whole … it is always part of a network of statements' . 217 Future research could expand my treatment of the 'conceptual archive' , by considering its wider 'network' . Research into non-elite discourses could build understanding of the archive's relational evolution.
These are paths for developing contributions made here: clarifying the conceptual archive's analytical scope, and expanding its parameters. In introducing the 'conceptual archive' , this article has illuminated possibilities for maturing security studies' sensitivity to past-present interpellation. I look forward to reading research realising these possibilities in future.