Available via license: CC BY-NC 4.0
Content may be subject to copyright.
Engaging an infrastructure of time
production with international law
Geoff Gordon*
This article explores international legal entanglements with the infrastructure
of Coordinated Universal Time (UTC), a socio-technical construction that com-
prises the globally definitive measure of seconds, minutes and hours working
in lockstep around the world. UTC is produced and maintained with
international law, through international administration with social, political,
material and technical dimensions. In analysing mutual entanglements of inter-
national law with infrastructural relations organized around UTC, the article
assesses their imbrication in globalized governance routines and distributive
patterns, instantiated here with reference to the field of finance. The article
further considers ways to engage legal-infrastructural relations for material
change in patterns of governance.
There are many sorts of time, and many ways of knowing and marking time to
do different sorts of work. This article explores international legal entangle-
ments with the infrastructure of a particular measure of time, namely
Coordinated Universal Time (or UTC). UTC comprises the globally definitive
measure of seconds, minutes and hours working in lockstep around the world.
Other time scales notwithstanding, UTC represents a hegemonic time scale,
produced and maintained by international law through international adminis-
tration with social, political, material and technical dimensions. In short, UTC
is a socio-technical construction.
1
Its scope is global, but the work it does is
*Senior Research at the Asser Institute. Many people have generously given their time to this manu-
script. I would like to extend my thanks to all of them. A partial list includes: Matilda Arvidsson,
Kevin Birth, Lianne Boer, Amy Cohen, Delphine Dogot, Isabel Feichtner, Daniela Gandorfer, John
Haskell, Joris van Hoboken, Andrea Leiter, Gregoire Mallard, Dimitri van den Meerssche, Niels
ten Oever, Sofia Stolk, Gavin Sullivan, Renske Vos, Wouter Werner, Kangle Zhang, and the organ-
izers of the InfraReg conference at NYU. I could not have asked for a better editor for this manu-
script than Sara Kendall, my thanks to her and the other editors at the London Review. Remaining
shortcomings are testament to my stubbornness. All URLs accessed 29 November 2021.
1 Cf. S Jasanoff and S-H Kim (eds), Dreamscapes of Modernity: Sociotechnical Imaginaries and the
Fabrication of Power (University of Chicago Press 2015).
London Review of International Law, Volume 9, Issue 3, 2021, 319–349
https://doi.org/10.1093/lril/lrab021
V
CThe Author(s) 2022. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial
License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commerc ial re-use, distribution, and
reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact
journals.permissions@oup.com
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
varied and particular. In this article, I will take up the particular work that the
production of UTC does in support of ‘lifeworlds’ such as Karin Knorr Cetina
describes for global currency traders, focusing here on high frequency trading
in the field of finance.
2
But constitutive elements in an infrastructure for glo-
bally standardised time production do not make obviously productive targets
for legal intervention. One aim for this study is to engage technical sites where
the administration of standardised time is imbricated in governance routines
and distributive patterns that are global in scope.
Benedict Kingsbury has pointed out that ‘infrastructural thinking [brings
together] the technical (the designed and engineered physical and software ele-
ments), the social (the human and non-human actants in their intricate rela-
tions), and the organisational (the forms of entity, regulatory arrangements,
financing, inspection, governance, etc.).’
3
Legal practice is implicated in each
of these dimensions, and in reciprocal fashion: ‘Law may intervene in the tech-
nical, the social and the organisational, and each of these may be embedded in
a particular environment of legal forms and relations.’
4
Paul Edwards, para-
phrasing Langdon Winner, describes the action of infrastructures, which he
calls law-like: ‘infrastructures act like laws. They create both opportunities and
limits; they promote some interests at the expense of others.’
5
And Brian
Larkin, addressing their governance dimensions, holds that infrastructures can
be reverse engineered, at least in theory, to ‘reveal forms of political rationality
that underlie technological projects and which give rise to an “apparatus of
governmentality”’.
6
I build upon parts of both quotes, Edwards’ and Larkin’s,
while also pushing back on a similar aspect in each. Edwards helpfully makes
plain the opportunities and limits produced by infrastructures, the way they
can promote some interests at the expense of others, which will be a primary
point of attention in the second and third sections following this introduction.
2 K Knorr Cetina and U Bruegger, ‘Inhabiting Technology: The Global Lifeform of Financial
Markets’ (2002) 50 Current Sociology 389.
3 B Kingsbury, ‘Infrastructure and InfraReg: On Rousing the International Law “Wizards of Is”’
(2019) 8 Cambridge International Law Journal 171, 179. Kingsbury’s suggestion to take up ’infra-
structural thinking’ takes special urgency in the context of the emergence of infrastructural
assemblages such as described by Gavin Sullivan in the context of global security practices. G
Sullivan, The Law of the List: UN counterterrorism sanctions and the politics of global security law
(Cambridge University Press 2020).
4 ibid.
5 P Edwards, ‘Infrastructure and Modernity: Force, Time, and Social Organization in the History of
Sociotechnical Systems’ in P Brey, A Rip, and A Feenberg (eds), Modernity and Technology (MIT
Press 2002) 185, 191.
6 B Larkin, ‘The Politics and Poetics of Infrastructure’ (2013) 42 Annual Review of Anthropology 327,
328.
320 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
But I want to bracket the ‘law-like’ ascription, because I mean to observe law
as part of the infrastructure, not something that exists independently of it.
Likewise, Larkin’s quote helpfully points governance towards issues of govern-
mentality, especially insofar as governmentality stands for governmental ra-
tionality, in which ‘the focal question of politics is ... not so much the
justification of state action as the governability of the social’.
7
Governmentality in this mode comprises what it takes ‘for an individual, and
for a society or population of individuals, to be governed or governable’.
8
These issues of governance and governmentality will be a concern throughout
this article. But I want to bracket the notion that an infrastructure can be re-
verse engineered to arrive at a governmentality that precedes it. I am interested
instead in the ways in which governmentality associated with an infrastruc-
ture—like Edwards’ ‘law-like’ effects—is produced through and with the infra-
structure at any given moment.
Like the time that it produces, the infrastructure for globally standardised
time is a particular construction: one infrastructural assemblage for one class
of time measurement. Even so limited, it can be put to work for innumerable
purposes—but still it cannot and does not do just anything. It supports a par-
ticular timescale to do particular work, distinct from other possible timescale
artefacts such as sundials, incense sticks, seasons, cigarettes, etc. As I will dem-
onstrate in the three sections following this introduction, it has been built to
do the work that it does based upon specific material and conceptual possibil-
ities, developed on the basis of specific interests, and mobilised for specific
purposes.
I have already gestured broadly at interests and purposes by reference to
the ‘lifeworlds’ that Knorr Cetina has described with respect to the foreign ex-
change market, materialising out of the socio-technical conditions that enable
time-based association in the field of global finance.
9
Those lifeworlds are con-
stituted with—and contribute to the construction of—a flow architecture, a
contemporary facet of the infrastructure for globally standardised time com-
prising persons, screens, clocks and information (among other things).
10
The
7 C Gordon, ‘Governmental Rationality: An Introduction’, in G Burchell, C Gordon and P Miller
(eds), The Foucault Effect: Studies in Governmentality (University of Chicago Press 1991) 1, 34. In
this Foucauldian register, governance refers to the conduct of conduct, by means of interventions
in social behaviour, whereas governmentality refers roughly to a framework by which government
comprehends one mode or rationale of governance to be preferable to another.
8 Ibid 36.
9 Knorr Cetina and Bruegger (n 2) 389.
10 K Knorr Cetina, ‘From Pipes to Scopes: The Flow Architecture of Financial Markets’ (2003) 4
Distinktion: Scandinavian Journal of Social Theory 7.
London Review of International Law Volume 9, Issue 3, 2021 321
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
flow architecture described by Knorr Cetina constantly produces and reprodu-
ces, moment to moment, a temporalised economic reality, and thereby privi-
leges a rationality of emergence that is capable of colonising other social
domains, as I will address in the second section following this introduction.
Among the lifeworlds that it supports, however, the interests and purposes at play
are not always coordinated or even complementary. Moreover, the infrastructure
for globally standardised time and the units of time that it produces are not
closed, fixed artefacts. Though the material assemblage of this infrastructure is
durable, it is also always changing. As I will demonstrate in the next three sec-
tions, the infrastructure for globally standardised time is contingent on constant
and deliberate support and intervention: its operation and effects are inscribed
with investments in (always ongoing) maintenance and support. These invest-
ments are steeped in law.
The Greenwich Meridian was once the site and product of an intense
contest under international law to determine a global index for time;
11
today
the Greenwich Meridian is a relic, with little relevance to a global administra-
tion of time, also established under international law, which relies on and reg-
ulates iterative algorithmic sampling of hundreds of nuclear clocks kept
around the world (the next section). I will look in the final section at examples
of change associated with innovations in time control clocks. Such network
investments in changing time technologies are driven by material interests and
possibilities. These material interests and possibilities, however, are competi-
tive in nature and not perfectly stable. The assemblage has been and remains a
site of competing material interests to determine or capitalise on possibilities
afforded by the infrastructure, and observing material sites of active invest-
ment indicates the stakes of globally standardised time production, as well as
the interests in play. A securities trader capable of producing a timestamp to
an extra decimal point, for example, can evade regulatory constraints by acting
after a deadline but rounding to remain within it.
12
This requires investment,
and pushes the infrastructure to propagate the next decimal point in its pro-
duction and distribution of standardised temporal units, with knock-on effects
reaching also to actors incapable of exploiting the additional decimal point.
For another example, the United States Department of Defense has committed
large sums of money to develop stable, miniaturised nuclear clocks for military
11 ‘International Conference Held at Washington for the Purpose of Fixing a Prime Meridian and a
Universal Day: Protocols of the Proceedings’ (October, 1884) <http://www.gutenberg.org/files/
17759/17759-h/17759-h.htm#Page_13>.
12 K Birth, ‘While the West Sleeps: Deglobed Globalization and Its Consequences’ in P Huebner et al
(eds), Time, Globalization and Human Experience (Routledge 2016) 109, 114.
322 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
deployment and strategic advantage.
13
Such clocks change conditions of secur-
ity, accessibility and mobility of authoritative measurements of standardised
time, and alter the mix of other assemblages in which they are imbricated.
In sum, the reproduction of the infrastructure for globally standardised
time is driven by competitive distributive stakes. Competition, however, has
not undone the construction, and here is one of the decisive contributions by
contemporary legal practice to the ongoing construction of the infrastructure.
As I will describe in the next section, legal practice stabilises competing inter-
ests around a consistent administrative architecture, enabling the sustained as-
semblage of a coherent infrastructure with consistent outputs despite
competitive conditions. Competition to exploit the next decimal point, for in-
stance, is enfolded with regulatory negotiations over the mix of clocks, algo-
rithms and protocols in the overall assemblage for standardised time
production under legal practice. Similarly, legal practice works to stabilise
expectations, or coordinate expectation horizons within the assemblage,
14
which will condition behaviour in any given site of activity—such as, for ex-
ample, with the shared understanding that a specific commodity will not be
readily transferable on the open market without good title (which will exhibit
an orderly transfer of ownership recorded in linear time).
The relationship is reciprocal: legal practice stabilises the assemblage des-
pite conditions of competition, and standardised time production stabilises the
intelligibility of some legal practices, despite pressures associated with expansive
economic conditions. Indexing ostensibly noneconomic events to standardised
time facilitates their legibility for transactional purposes (the second section
below), triggering a host of legal routines. Both time production and legal prac-
tice help to stabilise the hegemonic intelligibility of the other, including the par-
ticular sorts of expectations that may be recognised and communicated in
practice. Thus, a related role for international law is to coordinate expectation
horizons at the margins of networks, and from that vantage, to facilitate nego-
tiations and translations of communications across bounded networks—such as
when expectations generated in one network, for instance finance, can influence
expectations in another system, for instance health care.
15
Iwillreturninthe
conclusion to law’s material contributions in the interoperation of multiple
infrastructures. Each of these roles listed here—stabilising competition and
13 DARPA News, ‘Reducing Tics in the Tocks of Atomic Clocks: If GPS Goes Down, More Stable Atomic
Clocks Could Save the Day’ CHIPS: The Department of the Navy’s Information Technology Magazine (29
December 2015) <https://www.doncio.navy.mil/CHIPS/ArticleDetails.aspx?id=7274>.
14 N Luhmann, Law as a Social System (Oxford University Press 2004).
15 On the idea of legal regimes as the outcome of agonistic negotiations, see D Kennedy, ‘The Stakes
of Law, or Hale and Foucault!’ (1991) 15 Legal Studies Forum 327.
London Review of International Law Volume 9, Issue 3, 2021 323
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
expectations, and working as a switching mechanism at points of interface—
relies on mechanisms of institutional regulation established under international
law. I turn to regulatory mechanisms in the following section.
MAKING TIME: PRODUCING UTC TODAY
Guaranteeing the authoritative production of globally standardised time
today, in the form of UTC, is the responsibility of the Bureau International
des Poids et Mesures (BIPM). That responsibility includes coordinating the
different ‘universal’ times produced in laboratories around the world (which
produce so-called ‘local realisations’), a responsibility taken over in 1987 from
the International Time Bureau (BIH), which was formed at the 1912
Conf
erence internationale de l’heure radiot
el
egraphique,
16
and made part of
the International Astronomical Union at the 1919 Constitutive Assembly of
the International Research Council (founded then by Belgium, Canada,
France, Great Britain, Greece, Japan, and the United States).
17
The BIPM,
however, calculates the guiding UTC standard on the basis of another time
standard, International Atomic Time (TAI). It calculates TAI by drawing on
local realisations from a pool of more than 400 atomic clocks, maintained by
more than 70 laboratories, or ‘timing centers’.
18
In each of these national labo-
ratories, a master clock takes readings from a number of atomic clocks to aver-
age out an accurate reading, though the readings for the individual clocks may
be differently weighted for the average, and the algorithms for that weighted
averaging calculation vary across the national laboratories. The BIPM, then,
also applies a weighted average to the variety of averaged readings that it
receives from the national laboratories. In other words, no one clock actually
tells TAI, which is the product of multiple algorithms that together calculate a
weighted average of weighted averages. After completing that process to pro-
duce TAI, the BIPM then converts from TAI to UTC by adjusting for leap sec-
onds, which are only accounted for with the latter.
19
Approximately one
month after the initial readings at the national level have been taken, the
BIPM will circulate the TAI and UTC that they helped to produce via a com-
munication called Circular T. Just as no one clock ever tells TAI, none tell
16 International Radiotelegraph Convention (signed 5 July 1912) [1913] UKTS 139, <http://handle.
itu.int/11.1004/020.1000/5.19.61.en.100>.
17 A Blaauw, History of the IAU (Springer 1994).
18 Bureau International des Poids et Mesures, ‘BIPM Annual Report on Time Activities’ (vol 13,
2018) 49, <https://webtai.bipm.org/ftp/pub/tai/annual-reports/bipm-annual-report/annual_re
port_2018.pdf>.
19 BIPM, ‘BIPM Technical Services: Time Metrology’, <https://www.bipm.org/en/time-metrology>.
324 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
UTC, either, at least not in the present. A clock may be shown to have told
UTC, but only after the fact, as a sort of confirmed coincidence, once the
BIPM’s adjusted TAI calculation is communicated with Circular T (turning
the title of the communication into a double entendre).
20
The UTC production process makes clear enough that the construction is
not made of clocks alone, but comprises a technological network that includes
people (e.g., the scientists and cleaning personnel who maintain the various
laboratories, together with the accountants and managers who determine the
time and resources available to them), their techniques (e.g., for maintaining
stables of atomic clocks, and the changing calculations for the weighted aver-
age times kept among them), and other things. Moreover, the people and tech-
niques at the highest levels of UTC policy-making are part of a complex
bureaucratic structure, reflecting a mix of political and technocratic dimen-
sions. The BIPM, which administers the technological network, is an intergov-
ernmental organization, established pursuant to the Metre Convention of
1875, with 58 states members to date.
21
Its assumption of responsibilities for-
merly held by the International Time Bureau, established pursuant to a differ-
ent treaty, has already been noted. In addition, the BIPM defers to the
Radiocommunications Sector of the International Telecommunications Union
(ITU-R), though ITU-R holds no formal power over the BIPM. The ITU-R,
however, is a UN agency, reflecting ‘universal’ membership.
22
Whereas the
BIPM is made up of (nationally-appointed) metrology experts, the ITU-R is
made up of political delegates. While the political delegates typically form
working parties with experts closely associated with the BIPM, still the ITU-R
represents political considerations in the construction of UTC. Kevin Birth,
for example, reports on the possibility that the ITU-R has capped the level of
influence that the United States’ clocks (maintained by the United States
Naval Observatory, or USNO) can wield, below what the algorithms otherwise
might provide, despite (or because of) the technological dominance of USNO
clocks.
23
In sum: there is tension between the experts, politicians, and
machines in the reproduction of UTC,
24
a tension that illustrates the politics
and political possibilities of UTC, to which I return in the conclusion.
20 Bureau International des Poids et Mesures, ‘FTP Server of the BIPM Time Department’ <https://
www.bipm.org/en/time-ftp>.
21 Convention du me`tre (signed 20 May, 1875) <https://www.bipm.org/en/metre-convention>.
22 More precisely, the ITU, the parent organization, is the UN agency, and the ITU-R its sub-agency.
23 K Birth, ‘Time Standards and Rhizomatic Imperialism’ in C McGranahan and JF Collins (eds),
Ethnographies of US Empire (Duke University Press 2018) 237.
24 T Quinn, From Artefacts to Atoms: The BIPM and the Search for Ultimate Measurement Standards
(Oxford University Press 2012) 316.
London Review of International Law Volume 9, Issue 3, 2021 325
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
In keeping with the embodied nature of time produced across a wide-
spread infrastructure, the production of UTC and distribution of UTC does
not stop with the BIPM and its coordinate institutions. The networks that
communicate and distribute UTC are part of the production process. The
communication of UTC contributes to its production, and so the infrastruc-
ture for globally standardised time in its productive capacity also comprises
the networks by which UTC is communicated and distributed. Once, some
readers will recall, these sites of communicative/productive authority would
have been identified with examples such as the telephone number, maintained
by the telephone company, that would announce an automated reading of the
time when called. Today, these sites include the times told by computing devi-
ces, including personal computers, smart phones, tablets, etc. To communicate
this time, each computing device counts on the reliability of the product of
the interactions administered by the BIPM. Moreover, the time that comput-
ing devices communicate is also an integral part of their overall operation, in
the form of timestamping. Computers rely on timestamps to communicate in-
telligibly, both internally as well as networked. This is because communication
among computing devices occurs by means of one or more commands: the
proper sequencing among commands—and the legibility of that sequencing—
is essential to their intelligibility. As a result, the deep integration of time and
widespread computing processes is now part of the infrastructure for globally
standardised time.
To sequence each next command or communication, every action of a
computer is stamped and identified with time, marking the moment of each
command’s articulation and execution. Moreover, this process occurs at mul-
tiple scales: a singular command within a singular program of a singular com-
puter is timestamped, a peer-to-peer communication is timestamped, an
email is timestamped, a blockchain hash function is timestamped, and so on.
A closed system of communication might be able to rely on its own time
source for its own purposes, but a complex multiplicity of systems calls for a
reliable, global source, which is the role assumed by UTC, as produced by the
BIPM in consultation with the ITU-R. Because UTC is only ever actually pro-
duced after the fact, however, its reproduction and operationalization in com-
puter networks entails still more layers of mediation. For computers and other
devices to apply digital timestamps that produce a consistent result across net-
works, and in so-called real time, they rely on synchronization among linked
machines, all running their nearest approximation of what UTC will have
been confirmed to be. That synchronization requires consistent protocols. The
dominant protocol for the purpose has been Network Time Protocol (NTP),
formulated roughly in 1981 by David Mills, a veteran of the US Department of
Defense’s Advanced Research Projects Agency (now known as DARPA).
326 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
Mills is one of the early architects of the Internet, about which he
states plainly: ‘Having the wrong time could completely destroy the
network.’
25
Communication among machines via the internet (or otherwise)
would not be possible as we know it without reliable temporal sequencing for
each piece of information to be transmitted. To enable that coordination,
NTP functions according to a hierarchy among linked computers, servers and
clocks. The most dominant servers and clocks belong to the USNO, but the
NTP hierarchy includes up to 15 levels of stratification, with multiple clocks
and servers at each level. NTP selects among the available linked and stratified
devices by means of an algorithm to ascertain the closest approximation of
UTC in the network. In other words, NTP is an algorithm that fashions a
weighted average out of an approximation of a time that itself is a weighted
average of weighted averages, the results of which will only be circulated after
the fact.
26
Linear it is not, but precise it is. It is able consistently to produce
time units with apparent systemic consistency of roughly within 10–100
nanoseconds.
27
NTP, however, is no longer the most accurate available protocol. A rela-
tively new protocol, called Precision Time Protocol (PTP), aspires to syn-
chronise time to within one nanosecond of accuracy. The Institute for
Electrical and Electronic Engineers (IEEE), a non-governmental body, issued
PTP in 1997. PTP achieves its greater accuracy on the basis of several distinc-
tions from NTP.
28
To summarise broadly the distinctions that are relevant
here, PTP applies best to systems with their own atomic clocks and a small
number of connections, featuring symmetrical architecture among them. This
has two consequences. First, PTP tends to prioritise the time distributed by
the USNO, thus giving greater weight to US time technologies (apparently at
odds with the ITU-R). Second, PTP can be extraordinarily expensive, especial-
ly if it is applied in a network of any size or geographic scope, where a closed,
reliable and symmetrical loop of cables and connections becomes a prohibi-
tively difficult feat of engineering.
29
As a result, it is only deployed in situations
25 D Kukich, ‘Where Timing is Everything’ 11 (2002) University of Delaware Messenger <http://
www1.udel.edu/PR/Messenger/02/1/where.html>. See also D Mills, ‘“A Maze of Twisty, Turney
Passages”—Routing in the Internet Swamp (and Other Adventures)’ (18 May 2005) <https://
www.eecis.udel.edu/mills/database/brief/goat/goat.pdf>.
26 Birth (n 12) 109, 114.
27 D Mills, ‘NTP Precision Time Synchronization’ (5 July 2008) <https://www.eecis.udel.edu/mills/
database/brief/precise/precise.pdf>.
28 National Institute of Standards and Technology, ‘IEEE 1588 Standard for a Precision Clock
Synchronization Protocol for Networked Measurement and Control Systems’ (26 January 2021)
<https://www.nist.gov/el/intelligent-systems-division-73500/ieee-1588>.
29 Birth (n 12) 114–15.
London Review of International Law Volume 9, Issue 3, 2021 327
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
where it is worth the enormous expenditure, and in those cases, it is only
deployed by those who can afford it. This reality drives home the point
emphasised earlier: that the infrastructure for globally standardised time is a
particular construction, with particular interests invested and embedded in its
material build. With respect to the cutting edge of time-keeping technologies
today, there are a limited number of practices in which PTP’s level of accuracy
is worth the expenditure, which include select data mining operations and
what is known as fintech.
I turn next to fintech practices of algorithmic trading. Before I do, how-
ever, consider the production of UTC in comparison with both its imaginary
predicates and ultimate outputs. Its imaginary predicates describe an ideal linear
procession of identical units progressing simultaneously and in lockstep; its ul-
timate output is a practical and precise approximation of the same. The produc-
tion of UTC, however, is quite different. Every machine involved in the
production of UTC tells a unique version of it, none of them operate in actual
lockstep, and simultaneity is impossible to attain. The production of UTC pro-
ceeds recursively rather than linearly, with what might be analogised to cro-
cheting, or a loop-and-stitch pattern: to determine the present, the BIPM
regularly loops back to tell past time, while computers stitch the future (the
point at which the authoritative time will retroactively be told) together with
the present by approximating each next loop-back. In short, the production
process matches neither its uniform, linear idealization, nor the commodified
version of that idealization, the artefactual output. I raise this disjuncture to
make clear that the complex architecture of the system cannot be explained by
natural or intuitive inevitability, and its internal inconsistencies discount any
singular truth in its operation—the infrastructure for globally standardised time
and the globally hegemonic standardised time that it produces do not exist on
any transcendent basis; rather, the infrastructure exists because it works. This
raises the simple question: what does it work for?
30
There is not one answer to
this question, but different answers will point to different investments in the as-
semblage for different purposes. The next section offers one example.
TRADING TIME: TIME IN FINANCE AND HIGH FREQUENCY
TRADING
The previous section addressed the ways in which institutions of international
law, including the BIPM and ITU-R, regulate the infrastructure for globally
30 Cf. G Gordon, ‘Transnational Time’ in L Boer and S Stolk (eds), Backstage Practices of
Transnational Law (Routledge 2019).
328 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
standardised time. This section addresses entanglements of this infrastructure
with the world of finance, including the sort of work that the two do for and
with one other, as well as how legal practices are implicated in their interac-
tions. I focus on activities known as high frequency trading, an application of
algorithmic trading. Algorithmic trading today relies on artificial intelligence
to process enormous amounts of data to leverage information and trade for
profit. In the case of high frequency trading, algorithmic calculation is used to
make rapid and constant trades with massive sums, to generate profits from
what are typically small movements in financial devices such as commodities,
securities, and derivatives. Once controversial, these practices have become
routine, whereby money is ‘made’ or ‘lost’ according to a race among compet-
ing analyses and bids measured in nanoseconds but involving massive sums.
31
I focus here on high frequency trading because it allows me to foreground
the dimension of speed and, with it, matters of time, but it is important to
keep in mind that the production (and product) of time is deeply implicated
in financial practices beyond quantitative finance.
32
High frequency algorith-
mic trading is possible by virtue of a mutually implicating relationship be-
tween finance and time as that relationship is and has been materially,
technically and legally constituted—it is endogenous, at least in part, to each
of these vectors of practice, not some external in(ter)vention with respect to
any of them.
33
A fundamental connection between finance, time and technol-
ogy is clear already in volumes two and three of Marx’s Capital, which describe
the tendency towards increasing volume and speed in the circulation of fi-
nance capital as a systemic feature of capitalism.
34
In contemporary practice,
finance is constituted in part out of trade in temporal units, according to prac-
tices enabled and conditioned by law. In this situation, UTC is both an object
and agent of financial speculation: it is exchanged as a commodity, as abstract
equivalents and objectified units of value, while simultaneously driving the
31 S Patterson and G Rogow, ‘What’s behind High-Frequency Trading’ Wall Street Journal (New
York, 1 August 2009) 44.
32 See, eg, E Gilbert, ‘Common Cents: Situating Money in Time and Place’ (2005) 34 Economy and
Society 357; K Knorr Cetina, ‘How are Global Markets Global? The Architecture of a Flow World’
in K Knorr Cetina and A Preda (eds), The Sociology of Financial Markets (Oxford University Press
2004) 38; M Pryke and J Allen, ‘Monetized Time-Space: Derivatives—Money’s “New Imaginary”?’
(2000) 29 Economy and Society 264.
33 F Pasquale, ‘Law’s Acceleration of Finance: Redefining the Problem of High-Frequency Trading’
(2014) 36 Cardozo Law Review 2085.
34 Karl Marx, Capital: Volume II (2nd edn, 1893) <https://www.marxists.org/archive/marx/works/
1885-c2/index.htm>and Capital: Volume III (2nd edn, 1894) <https://www.marxists.org/archive/
marx/works/1894-c3/index.htm>.
London Review of International Law Volume 9, Issue 3, 2021 329
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
technologies that perform the exchange of commodified value and make it
legible.
The mutual implication of time, technology, finance and law has exhib-
ited a particularly visible character since the early 1970s, when the demise of
Bretton Woods portended the ascent of speculative finance and the increasing
commodification of debt.
35
Under Bretton Woods, national currencies were
maintained in a fixed price relationship to one another. With the collapse of
Bretton Woods, the prices of national currencies were allowed to ‘float’, or
vary up or down relative to one another, as with other commodities. In the
post-Bretton Woods landscape, the market for exchanging foreign currencies
led the ascent of speculative finance on a global scale.
36
At the same time, na-
tional and international regulatory initiatives provided banks with more ways
and means to sell debt, expanding the commercial market for debt products as
well as products based on the assignment or expectation of transactions and
changes in designated values at designated times.
37
In a relatively small
amount of time, the amount of debt circulating in financial markets, in the
form of bets on future behaviour (of whatever sort, monetised for market pur-
poses), outstripped by far the amounts of liquid currency and domestic prod-
uct.
38
By the International Monetary Fund’s (IMF) measurements, private side
debt (so-called) has tripled since 1950, and government debt, which had been
in decline until the mid-1970s, has gone up sharply ever since.
39
As a result,
the number of temporally-indexed objects and instruments of speculative fi-
nancial interest multiplied. The rise of temporally-indexed speculative instru-
ments brought with it a new practice community in the economic world of
35 S Strange, Casino Capitalism (Manchester University Press 1997); S Strange, ‘The New World of
Debt’ (1998) 230 New Left Review 91.
36 Knorr Cetina (n 10).
37 A Campbell and E Bakir, ‘The Pre-1980 Roots of Neoliberal Financial Deregulation’ (2012) 46
Journal of Economic Issues 531; R Felder, ‘From Bretton Woods to Neoliberal Reforms:
International Financial Institutions and American Power’ in L Panitch and M Konings (eds),
American Empire and the Political Economy of Global Finance (Palgrave 2008) 175.
38 Cf. G Teubner, ‘A Constitutional Moment? The Logics of “Hitting the Bottom”’ in P Kjaer, G
Teubner and A Febbrajo (eds), The Financial Crisis in Constitutional Perspective: The Dark Side of
Functional Differentiation (Bloomsbury 2011) 3–4.
39 S Mbaye and M Moreno Badia, ‘New Data on Global Debt’ (IMFBlog: New Insights on Economics
and Finance, 2 January 2019) <https://blogs.imf.org/2019/01/02/new-data-on-global-debt/>. The
IMF’s Global Debt Database, from which these numbers are taken, is available online: IMF,
‘Global Debt Database’ <https://www.imf.org/external/datamapper/datasets/GDD>. Cf. S Mbaye,
M Moreno Badia and K Chae, ‘Global Debt Database: Methodology and Sources’ (IMF Working
Paper No 18/111, 14 May 2018) <https://www.imf.org/-/media/Files/Publications/WP/2018/
wp18111.ashx>. The World Bank documents three waves of growth in debt since 1970, highlight-
ing especially marked growth since 1980. M Kose, P Nagle, F Ohnsorge and N Sugawara, Global
Waves of Debt (World Bank Group 2020).
330 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
investment, and new ‘lifeworlds’, in the language of Knorr Cetina. Those life-
worlds are temporal ones, constituted by roughly simultaneous engagement in
the constant flux of speculative market prices.
40
They correspond with emer-
gent governmental rationalities that exhibit a tendency to colonise other social
domains, as I will discuss below.
After the demise of Bretton Woods, the loosening regulatory environ-
ment and multiplying ways of extracting profits from changes over time cre-
ated the conditions for what Susan Strange called ‘casino capitalism’,
emphasising the element of speculation.
41
Three aspects of the changes that
followed the collapse of Bretton Woods are pertinent here. For one, the specu-
lative movement of information and capital have created temporally-defined
economic lifeworlds of global scope, exhibiting what Knorr Cetina describes as
a ‘flow architecture’. In that flow architecture, scopic technologies allowing
the observation and projection of information combine with time measure-
ments to enable economic practices attuned to the moment, as I will further
describe below. Two, the flow architecture has enabled speculative economic
activities to reach and condition what had been considered non-economic
domains—this is the colonial capacity to which I have alluded, and to which I
will return. Finally, the common unit for exchange or special currency moving
within this flow architecture is time.
As a unit of exchange enabling diversification of potential vehicles for
debt, time has been widely observed to enable debt as an instrument of specu-
lation delinked from production.
42
Derivatives are illustrative, representing
bets on the value of a commodity or commodified index at a designated mo-
ment, in which the point is neither production nor possession. The investor in
derivatives rarely if ever engages in production or takes any possession.
43
Only
the exchange value of the chosen index at the programmed time counts.
Barbara Adam makes clear the central value of time to these post-Bretton
Woods economic practices: ‘most money is made on the speculative market
that trades not with goods but with time. This involves bets on future prices of
the stock market, currency prices, interest rates, even on the entire stock mar-
ket indices. The trade in time [in 1995] equal[ed] the entire stock of
40 Knorr Cetina and Bruegger (n 2).
41 Strange (n 35).
42 W Hope, ‘Time, Communication and Financial Collapse’ (2010) 4 International Journal of
Communication 649, 650–2; W Hope, ‘Global Capitalism and the Critique of Real Time’ (2006) 15
Time and Society 275, 276–7.
43 D Bryan and M Rafferty, ‘Financial Derivatives as Social Policy Beyond Crisis’ (2014) 48 Sociology
887.
London Review of International Law Volume 9, Issue 3, 2021 331
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
productive fixed capital of the world’, and it has grown since.
44
By 2018, the
International Monetary Fund estimated that global debt, including both ‘gov-
ernment and private sides of borrowing’, totalled $184 trillion, outstripping
global GDP by 225 per cent.
45
This is wealth (and debt) accumulation on the
basis of wagers indexed to time.
46
Debt instruments enable trading in time for
economic profit. A unitary time scale is necessary for the purpose: there must
be an agreed set of temporal markers—and agreed means to generate and con-
firm those markers—to manage expectations under law. This will have ‘down-
stream’ temporal consequences for the object of the bet, including any
production and distribution processes involved, which will now be linked by
its financing to the time scheme of the speculative instrument such as the de-
rivative.
47
Downstream consequences reach all the way to the bodies of work-
ers, obligated to labour according to a time and pace determined by massive
and fast-moving flows of speculative debt.
48
The trade in time also has consequences that spread out horizontally. If
one consequence is downstream effects on the object of speculation, including
the bodies of workers occupied with it, this second consequence is the cross-
stream effect, whereby multiple potential objects of speculation are compara-
tively assessed for the relative profit they might yield once translated into com-
modified temporal terms.
49
The potential return on each speculative wager
can be notionally measured against the potential return of any other specula-
tive possibility intelligible to potential speculator-investors. This is one way in
which speculative economic activities have reached and conditioned what had
been considered non-economic domains, by indexing them to common tem-
poral markers under law for speculative economic purposes. The only ‘hori-
zontal’ limit on trade in time concerns the intelligibility of the potential
wager—can it intelligibly (and so actionably) be translated as an object of
speculation? The language of intelligibility thus begins with temporal markers
44 B Adam, ‘The Gendered Time Politics of Globalization: Of Shadowlands and Elusive Justice’
(2002) 70 Feminist Review 3, 22.
45 Mbaye and Moreno Badia (n 39); IMF (n 39).
46 Cf. M Feher, Rated Agency: Investee Politics in a Speculative Age (Zone Books 2018).
47 Cf. K Do¨rry, ‘Strategic Nodes in Investment Fund Global Production Networks: The Example of
the Financial Centre Luxembourg’ (2015) 15 Journal of Economic Geography 797; D MacKinnon,
‘Beyond Strategic Coupling: Reassessing the Firm-Region Nexus in Global Production Networks’
(2012) 12 Journal of Economic Geography 227.
48 I cabin here for reasons of space the biopolitical dimensions of the infrastructure for globally
standardized time, to focus within limited space on select socio-technical dimensions of its con-
struction, but it must be clear that its inscription on and in bodies is a material part of the overall
assemblage, not some mere externality.
49 Feher (n 46).
332 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
to make dynamics of duration and differentiation amenable to speculation. An
infrastructure for globally standardised time distributes general intelligibility
to enable the expansive and colonising speculative possibility. It incorporates a
combination of regulatory, technical and material components to ensure com-
parability, consistency and predictability across speculative transactions.
50
Here the role of temporal units not just as objects but also as agents is appar-
ent: temporal units here are not solely units of currency for speculation, but
are instrumental in the construction of the legibility that makes speculative
transactions possible. In this socio-technical system, every act and transmis-
sion must be digitally timestamped. Birth describes the importance of the
timestamp:
Global capitalism relies on accurate timestamps. All messages include
a timestamp generated by the exchange on which the order is posted,
and a collection-point timestamp where the order could be executed.
For a traded security, every time its price is adjusted, it receives a new
time stamp. Some orders also have expiration times. Together, these
collected times constitute what is called ‘transaction time’.
Transaction time does not represent a moment in time, but a duration
of a piece of data in a database—in this case, an offer in an exchange’s
trading sequence.
The timestamps that constitute transaction time allow the represen-
tation of the sequence of activity pertaining to each order and offered
security. This is important because trading volumes are greater than
servers can process in real time, so the transaction time system charts
the sequence of orders, quotes, and price changes. The rapid global
circulation of capital then unfolds in two times—the real time as expe-
rienced by people, and a representational time as generated by the
sequencing of timestamps.
51
The two times that Birth describes are evident in what I called the ‘down-
stream’ effects, in which the transaction time for capital (including the ficti-
tious capital of leveraged debt) will outpace and so pressurise the time on the
job as experienced by workers in the labour of production.
Both the crucial role of the timestamp and the disjuncture from time as
experienced by people are especially clear in the example of algorithmic trad-
ing practices today. Algorithmic trading includes bulk trades in stocks and
other financial instruments as a bet on movement in their short-term values.
50 M Pryke and J Allen, ‘Monetized Time-Space: Derivatives—Money’s “New Imaginary”?’ (2000) 29
Economy and Society 264.
51 Birth (n 12) 115–16.
London Review of International Law Volume 9, Issue 3, 2021 333
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
In the case of high frequency trading, agents typically unwind every position
by the end of every 24-hour cycle, thus never maintaining an ownership inter-
est for long. To turn a profit on such short-term movement, however, algo-
rithmic, high frequency trading involves massive purchases or sales in
constant succession on the order of nanoseconds. The practice also includes a
vast number of trades that do not (and are not meant to) happen, offers that
are made and revoked nearly instantaneously to game the market. The com-
bination of extreme speed, precision and repetition is crucial for these pur-
poses.
52
Further, these attributes are also indicative of the flow architecture
that Knorr Cetina describes. Knorr Cetina’s flow architecture includes a dis-
continuity ‘between the spatial or physical world we usually conceive of, and
that of a timeworld.’
53
Whereas the spatial world as Knorr Cetina describes it
is mapped and made navigable with symbolic structures that endure, the time-
world established by flow is distinct and more immediate: ‘In a timeworld or
flowworld ... the content itself is processual—a “melt” of material that is con-
stantly in flux, and that exists only as it is being projected forward and calls
forth participants’ reactions and contribution to the flux.’
54
The flux relies on
scopic technology, which Knorr Cetina describes as ‘reflexive mechanisms of
observation and projection.’
55
Scopic technology is also information technol-
ogy, which further explains the phenomenon of flow: ‘the material on [the fi-
nancial practitioner’s] screen can disclose itself as information only in as far as
it is new compared to earlier material. The new is “presenced” as-things-
happen and vanishes from the screens as newer things come to pass. This sort
of reality is inherently temporal, which is what [is also meant by] “flow”.’
56
Information managed according to a specific regime of time measurement and
distribution enables the melt and flux that gives casino capitalism its radical
immediacy and expansive potential. The faster the succession of new informa-
tion, the more ‘liquid’—the fuller and more dynamic—the flow.
Thus, the fuller and more dynamic the flow of temporalised information,
the greater the colonising capacity to subject seemingly unrelated social
domains in the spatial world to the emergent governmentality of the financial
lifeworld. The growth in that capacity is charted by the rising volume and
52 M Chlistalla, B Speyer, S Kaiser, and T Mayer, ‘High-Frequency Trading’ (2011) 7 Deutsche Bank
Research <https://secure.fia.org/ptg-downloads/DBonHFT2-11.pdf>.
53 Knorr Cetina, (n 32) 39.
54 ibid 40.
55 ibid.
56 ibid 43.
334 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
variety of debt instruments since the demise of Bretton Woods.
57
Contemporary time technologies, including timestamps and nuclear clocks,
today combine to enable a constant succession of volumes of information in
nanoseconds, producing flows of financialized information exhibiting ex-
treme, self-reflexive dynamism. The ultimate limits on that movement of in-
formation and capital in algorithmic trading markets are technological limits
on the communication of timestamped computations. Trading firms deploy
algorithms to mine data, rapidly processing an enormous and constantly
changing mix of information, organised by timestamp. The same algorithms
then automatically produce bets on the products of their data mining activity,
the bets registered by timestamp technologies capable of sequencing transac-
tions at astronomical speeds.
58
Returning to a point raised above, the cost and
sophistication of the technologies that are necessary to this practice become a
sort of gatekeeper. Only those who can afford it—whether individuals, firms,
governmental entities, etc.—invest actively and substantially in the design and
operation, and then only to ends perceived by some metric to offset the costs.
As Birth puts it, ‘[a]ccess to timestamp precision becomes a subtle creator of
stratification embedded in the production of timestamps’.
59
Because algorithmic trading occurs in competitive markets, the twin
requirements of speed and precision become the subject of intense techno-
logical competition. Competing algorithms process countless data points for
patterns in the market value of any given equity or commodity. The data
points can come from popular media, specialist databases, insider communi-
cations, and so on: the ‘winning’ algorithm will be the fastest to access and sift
enough information to approximate closely the movement in price—which
movement will include the effect of the winning bid, as well as the effects of
the countless unsuccessful bids, including unsuccessful bids that are meant to
game the price, rather than to win the wager.
60
In short, the financial practices
to predict prices according to temporal markers also co-produce them by
means of the same practices and technologies.
61
The result is orders placed
and revoked in nanoseconds, with positions typically not lasting much longer
57 M Lazzarato, The Making of the Indebted Man (Semiotext(e) 2011).
58 Chlistalla et al (n 52).
59 Birth (n 12) 115.
60 Cf. M Mazzucato, The Value of Everything: Making and Taking in the Global Economy (Hachette
2018); and A Cartea and J Penalva, ‘Where is the Value in High Frequency Trading?’ (2012) 2
Quarterly Journal of Finance.
61 For extended insights in this vein, see D MacKenzie, F Muniesa, and L Siu (eds), Do Economists
Make Markets? On the Performativity of Economics (Princeton University Press 2007).
London Review of International Law Volume 9, Issue 3, 2021 335
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
than that, representing vast sums of money in constant speculative movement.
That movement charts a fluid stream of values that condition distinct, down-
stream value-producing practices, and have cross-stream repercussions as well.
Contemporary time technologies play a central role by enabling vastly acceler-
ated and expanded practices for potential wealth accumulation by
speculation.
62
In this new material environment, however, the notion of finance as an
enterprise and the regulatory regime adapted to it have been destabilised.
63
The speeds involved are approaching a hard limiting condition, namely the
speed of light, which places special stresses on regulatory mechanisms and sys-
temic coherence generally: the horizon of relativity (at light speed) challenges
systemic demands for consistency and coordination.
64
The stress is prompting
a radical reconsideration of time within the field of financial regulation. James
Angel, reviewing the coordination challenge from a regulator’s perspective,
has called to revisit and reconstruct basic intuitions of time for legal govern-
ance purposes.
65
Angel’s concern underscores the entanglement of financial
regulation with the conduct that it would conduct: ‘intuitions—and regula-
tions—gained from low-speed markets also need to be modified as trading
approaches the speed of light.’
66
The governmental posture here is closely
attuned to the dilemmas that market behaviour creates for market partici-
pants, and the ways in which mutual participation breaks down from use of
time technologies in practice. ‘As trading speed increases, physical issues and
lessons from physics become more important for regulators and practitioners
alike’ because ‘a market participant [will] not really know if a published “firm”
quote is actually alive (can be traded on) or dead without actually submitting
an order to trade against it.’
67
Pulling back the lens, Wayne Hope observes the overall operation of in-
formation and communications technology (ICT) behind algorithmic, or high
frequency trading capabilities:
62 L Adkins, The Time of Money (Stanford University Press 2018).
63 S Omarova, ‘Dealing with Disruption: Emerging Approaches to Fintech Regulation’ (2020) 61
Journal of Law and Policy 25; S Omarova, ‘New Tech v. New Deal: Fintech as a Systemic
Phenomenon’ (2019) 36 Yale Journal on Regulation 735; R Buckley, ‘Reconceptualizing the
Regulation of Global Finance’ (2016) 36 Oxford Journal of Legal Studies 242.
64 M Lenglet and J Mol, ‘Squaring the Speed of Light? Regulating Market Access in Algorithmic
Finance’ (2017) 45 Economy and Society 201.
65 JJ Angel, ‘When Finance Meets Physics: The Impact of the Speed of Light on Financial Markets
and Their Regulation’ (2014) 49 Financial Review 272.
66 ibid 272.
67 ibid 279, 276.
336 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
ICTs generally have become the nervous system of globalizing capital.
Thus, computers, microchips, satellites and cable networks shape the
reach, velocity and supervening power of global finance. As govern-
mental control over exchange rates and capital movements dimin-
ished, private financial institutions were able to generate multiple
units of credit and currency ... Special satellite and internet linkups
have enabled stockbrokers, institutional traders and personal investors
to buy and sell shares concurrently in different stock exchanges.
Barbara Adam’s observations about ICT time are pertinent here; the
globally networked drive towards instantaneity and simultaneity is a
constitutive feature of financial profit making ...
68
The technological regime that Hope describes is not a neutral one. It
works better for and with some interests than others. Only a select, capitalised
few can compete for the profits that it yields. Among the many who are not so
privileged, labour suffers on multiple fronts: the infrastructure favours the
movement of temporalised information and things that can move in the form
of information (such as money and capital) in a way that it does not privilege
the movements of labour (in multiple senses, including the movements
required in physical labour and the movement of workers in pursuit of jobs
dislocated pursuant to financial mandates). The disadvantage is not arbitrary.
The temporal infrastructure predicated on UTC has been built into an espe-
cially profitable partner for those capable of harnessing ICT power. Finance
capital, consisting of transactions in time-based information delinked from ac-
tual goods, is one example. Thus, finance is an exemplary answer to the ques-
tion posed earlier: the infrastructure for globally standardised time works for
finance capital. For the same reason, finance capital is specially invested in
continued, competitive development of the infrastructure in order to master
self-reflexive markets in temporalised information, with particular distributive
consequences.
The ascendance of financial speculation and finance capital has also cre-
ated other conflicts, arguably opening up fissures within capitalism itself. The
sheer volume of finance capital circulating independent of production has cre-
ated a set of incentives and ends that are not entirely consonant with dynamics
traditionally associated with capitalist institutions. This is sometimes charac-
terised as a tension between the ‘finance economy’ and the ‘real economy’.
69
But the tension perhaps goes farther, exposing weaknesses in foundational
68 Hope, ‘Global Capitalism and the Critique of Real Time’ (n 42) 277.
69 Cf. S Hall, ‘Geographies of Money and Finance III: Financial Circuits and the “Real Economy”’
(2013) 37 Progress in Human Geography 285.
London Review of International Law Volume 9, Issue 3, 2021 337
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
economic notions, including the definition of value that yields measures such
as gross domestic product, or GDP.
70
The tension between modes of economic
practices, keyed to the use of time technologies, has also been described by het-
erodox and interdisciplinary economic observers as a potential source of hope
for transformation.
71
Among them, there is a specific current of hopeful cri-
tique that sees in the deployment of time as a speculative instrument of profit,
a technology that has overrun the ambitions that made it useful as an instru-
ment of capital in the first place. Robert Hassan, reviewing the high-speed
time techniques of financial networks, puts it this way: ‘In the name of
“efficiency”, neoliberalism has abrogated social control to both “market
forces” and computer networks of automation. [But e]ven those “in control”,
those in the boardrooms and cabinet offices of the great and powerful are es-
sentially “out of control”’.
72
Birth, in turn, refers to the consequences as part of a rhizomatic empire
of time. By rhizomatic, Birth means roughly that the consequences of deci-
sions by figures in cabinet offices and board rooms exceed the intent behind
them. This corresponds with Angel’s observation that the field of finance is
destabilised for regulators and traders alike by competitive deployment of time
technologies. The result, following Birth, has been a loss of directed, institu-
tional control: ‘The rhizomatic nature of this empire of time emphasises inter-
national coordination even as it diminishes the ability of any one interest to
control the empire. Yet, what is given up in terms of control is gained in the
ability to move information and capital quickly and audit such movements
precisely.’
73
Though the system of standardised time that Birth describes is not a
directed one and is not amenable to unitary control in any conventional sense,
it bears keeping in mind (pace Hassan) that limited classes of people continue
to benefit disproportionately from its operations in a field such as finance.
These observations underscore the stakes of law, to constitute new possibilities
for coordination and control under competitive conditions, for instance as
called for by Angel. They also point to new opportunities to engage the mater-
ial complexity of the infrastructure under conditions of unsettled design.
74
I
take this up further in conclusion. First, however, I turn to one technology
70 Mazzucato (n 60); J Gertner, ‘The Rise and Fall of the GDP’ New York Times (13 May 2010)
<https://www.nytimes.com/2010/05/16/magazine/16GDP-t.html>.
71 K Dorre, S Lessenich, and H Rosa, Sociology, Capitalism, Critique (Verso 2015).
72 R Hassan, ‘Network Time and the New Knowledge Epoch’ (2003) 12 Time and Society 226, 238.
73 Birth (n 23) 242.
74 Cf. D van den Meerssche, ‘International Organizations and the Performativity of Measuring States:
Discipline through Diagnosis’ (2018) 15 International Organizations Law Review 168, 174.
338 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
that has factored historically into that design, namely timestamping, as it was
first practiced on factory floors with analogue time control clocks. The de-
scription to this point of systemic characteristics of the infrastructure for glo-
bally standardised time risks giving the impression of an inevitable or
automatic technological development, but it is not that—though neither is it
arbitrary. The brief, relatively narrow historical detail that follows, concerning
one example of technological development, aims to demonstrate how the in-
frastructure is amenable to particular interventions that contribute to change
over time.
STAMPING TIME: THE TIME CONTROL CLOCK AND
MATERIAL INNOVATION IN THE INFRASTRUCTURE FOR
GLOBALLY STANDARDISED TIME
The development of the infrastructure for globally standardised time has not
been the product of black boxed intervention(s) from beyond. In the first sec-
tion, I examined the iterative and recursive structure of the legal-technical ad-
ministration of the infrastructure. In the previous section, I pointed to select
practices by which financial practices have been actively implicated in the de-
velopment of the infrastructure. In this section, I offer a historical example of
material innovations in the infrastructure. The example includes the mutual
implication of economic and legal-regulatory practices. Timestamps have al-
ready come up repeatedly, as they have played a role in each of the prior sec-
tions. I focus here on an analogue precursor to the digital timestamp.
Analogue timestamping is not the same technology as its contemporary digital
counterpart. The analogue timestamp is a mode of social and temporal control
associated with the factory floor and time control clocks rather than main-
frames and motherboards. But as will be clear, it was the pioneers in time con-
trol clocks and analogue timestamping who became pioneers in digital
timestamping. Analogue timestamping formed part of the technical imagin-
ation and matrix of possibility for the development of digital timestamping.
This history of timestamps begins with time control clocks, clocks that
synchronise and record events, originally to determine workers’ time on the
job by stamping a designated clock time (typically synchronised to some other
authoritative time source) on a card or paper to record presence and absence,
activity and inactivity. Charlie Chaplin spotlighted the practice in Modern
Times, in a scene in which Chaplin the worker is so routinised to time control
that he stops to ‘punch’ his time card while running from the cops.
75
The first
75 Modern Times (C Chaplin dir, 1936).
London Review of International Law Volume 9, Issue 3, 2021 339
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
influential analogue time control clock is generally credited to Willard le
Grand Bundy, and invented, patented and produced by the Bundy
Manufacturing Company in the late 1880s.
76
By 1900 Bundy was already ag-
gressively pushing operations overseas, including licensed production in
Germany for Bundy time clocks built in the Wu¨rttembergische Uhrenfabrik in
Schwenningen, and distributed by the Continentale Bundy-Gesellschaft,
founded in Stuttgart on 7 April 1900.
77
Bundy’s time clock utilised a key
mechanism: each worker was assigned a key that triggered wheels to stamp the
time on a paper tape.
Two other competitors at the time were Alexander Dey’s dial recorder,
and Daniel Cooper’s Rochester Recorder. Cooper invented the Rochester
Recorder in the late 1890s, then sold the patent and production rights to the
Willard and Frick Company. The Rochester Recorder worked by timecard,
with workers punching times of arrival and departure on a pre-printed card.
This became the iconic image of factory and work-place time discipline and
control. In 1900, Willard and Frick merged with the Bundy Manufacturing
Company, as well as another competitor, the Standard Time Stamp Company.
Following the merger, the company was aptly renamed the International Time
Recording Company, and given the motto: ‘Safeguarding the minute’. The
merger was orchestrated and overseen by George W Fairchild, who had joined
the Bundy Manufacturing Company in 1896 as an investor and director. From
1907 onwards, while still with the company, Fairchild also served as congress-
man for New York for six terms, sitting on the Committees of Foreign Affairs
and Ways and Means, becoming the ranking Republican on the latter and
thereby playing a substantial role financing the US war effort in World War
One.
78
Alexander Dey invented his machine in Glasgow. His brother, living in
the United States, had opened a company with him in 1893. Once sales took
off, Alexander joined his brother in the US in 1903, and by 1904 they had
founded the Dey Time Register Company. Their main clientele included rail-
ways and the Post Office, both involved in the propagation of standardised
time that would become UTC.
79
Dey’s dial recorder was named for its large
dial with workers’ numbers (in place of names) on it, to be selected by pointer.
76 IBM, ‘IBM Highlights, 1885–1969’ (2001) 2, <http://www-03.ibm.com/ibm/history/documents/
pdf/1885-1969.pdf>.
77 G Kopf, Zeit-Ordnung: Eine Geschichte der Stechuhr (PhD dissertation, Bauhaus-Universita¨t
Weimar, 2002) 19.
78 J Sullivan (ed.), The History of New York State (Lewis Historical Publishing 1927) 321–3.
79 G Gordon, ‘Railway Clocks’ in J Hohmann and D Joyce (eds), International Law’s Objects (Oxford
University Press 2018) 387.
340 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
A separate, smaller wheel had six markings, indicating arrival in the morning,
break for lunch, return at noon, quitting in the evening, extra time at work,
and extra time off. A worker selected the identification number on the dial,
the time-work status on the smaller wheel, and triggered a stamp on paper
kept within the machine. The stamp was marked in green or red ink, by means
of a two-toned ribbon patented by Dey, with green indicating punctuality and
regular working hours, and red indicating a late arrival, early departure, over-
time or time off. Repairs and maintenance were complex, but the paper record
was simple to read and efficient for accounting purposes, which made the ma-
chine popular.
80
In 1907, the International Time Recording Company pur-
chased the Dey Time Register Company. By this point, the International Time
Recording Company was far and away the dominant factory time control tech-
nology producer internationally. In 1911, following a series of further mergers,
the company was renamed again, becoming the Computer-Tabulating-
Recording Company. The Computer-Tabulating-Recording Company
changed its name once more, in 1924, to its current name, International
Business Machines, or IBM.
81
IBM, of course, has gone on to become a dom-
inant company in computing, as what began with time control clock innova-
tions developed through generational change into global computing and
business management interests.
The particular assemblage of the time control clock and the pervasive in-
dustrial practice of timestamping that it enabled contributed to change in the
valence of standardised time in daily life, as chronicled by E P Thompson, who
referred to practices of time control under industrial capitalism as time- and
work-discipline.
82
Time discipline entailed an entanglement of daily rhythms
that is already apparent on the face of Dey’s time wheel, marking the morning,
evening and lunch. Thompson described the development as ‘one of increas-
ing rationalization in the service of economic growth’.
83
Rationalization was
also a key term in the colonial expansion of globally standardised time, as part
of a rhetoric of science and civilization.
84
In the cotton mills of colonial India
and England alike, time control clocks imprinted the units of globally
80 Kopf (n 77) 21.
81 By 1958, IBM had expanded and sold off its Time Control Equipment division to the Simplex
Time Recorders Company. By that time, time control clocks had become ubiquitous in the control
of labour, and had begun to extend beyond the factory. And while IBM had divested from its roots
in time control technologies aimed at the factory floor, it had moved into the next generation of
timestamping technologies, inside computers: IBM (n 76) 5–6, 21.
82 EP Thompson, ‘Time, Work-Discipline, and Industrial Capitalism’ (1967) 38 Past and Present 56.
83 ibid.
84 G Gordon, ‘Imperial Standard Time’ (2018) 29 European Journal of International Law 1197.
London Review of International Law Volume 9, Issue 3, 2021 341
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
standardised time on factory work and factory workers, who carried it beyond
the shop floor, multiplying its reach alongside other technologies such as the
wristwatch.
85
The expansion was a bitterly contested one, occasioning strikes
and violence.
86
But legal regulation of the working day ultimately helped to
stabilise the contest around the competence of the time control clock and its
timestamps, until they became ubiquitous in still more working contexts,
including physical spaces from professional offices to motherboards. Since the
heyday of the analogue timestamp and the time control clock, time notations
printed with ink on paper have been succeeded by electronic processes to pro-
duce digital timestamp records, and the deployment of time as an instrument
of value extraction on the factory floor has expanded to countless other (digit-
al) contexts. In sum, the assemblage of the time control clock and the practice
of timestamping were enrolled into an expansive socio-economic program
associated with rationalization, imperial governance, colonial practices, and
legal-regulatory policy.
This brief history of factory time control clocks and timestamps illustrates
the sort of material qualities and contingencies that have gone into the devel-
opment of the infrastructure for globally standardised time. Some of the ma-
terial elements can be generalised: timestamping was developed on the basis of
making specific temporal information both mobile (transferable across people
and machines in the form of a paper record) and legible (readable on its face,
using numbers and colours for specific purposes). The relative ease of mobile
legibility facilitated the success of the technology as it was enrolled into expan-
sive economic programs, materially supporting the development of corporate
information management practices. On this socio-technical basis, what
Thompson called time discipline has continued to expand and evolve in ever
more expansive networks of people and things; the use of time discipline to
extract value in economic contexts has both expanded (in keeping with the
‘cross-stream’ movements described in the previvous section) and become
more intense (in keeping with the down-stream effects also described there).
Writing in the tradition of Thompson, David Harvey has famously described
the result as time-space compression.
87
Organizational innovations involving
mechanical stamps and paper, numbers, time and clock became predicate for
successive innovations.
85 D Mumby, Organizational Communication: A Critical Approach (Sage 2012) 59.
86 Cf. S Wolcott, ‘Strikes in Colonial India, 1921–1938’ (2008) 61 Industrial and Labor Relations
Review 460 (with thanks to Ntina Tzouvala for this reference).
87 D Harvey, The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change
(Blackwell 1989) 284.
342 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
There is a startling similarity in the role played by the analogue time-
stamp with that played by the price ticker and ticker tape in the formation of
the flow architecture observed in the foreign exchange market by Knorr
Cetina. The ticker tape, in which prices were inscribed on a running band of
paper as they were communicated in linear time, ‘temporalised the complexity
of earlier market transactions’.
88
By ‘sequentializ[ing] and display[ing] past
market activities’, Knorr Cetina argues, the ticker tape can be seen ‘as a first
step toward’ the temporal lifeworlds associated with computer-based scopic
systems.
89
The ticker ‘initiated a trend that pre-cast some aspects of scopic sys-
tems: it sequentialised the market, initiating a price-and-volume flow that be-
came at the same time an information flow.’
90
Like the ticker tape, the
analogue timestamp was precedent to an emergent temporal regime that
developed with the heightened speed and capacity enabled by digital technolo-
gies. The materials of the time control clock and timestamp together registered
data to order the workplace by regulating flow (of workers) with temporal in-
formation. In this capacity, the assemblage of the time control clock was
enrolled into the industrial program of the factory, and contributed to its
becoming. In turn, the time control clock as an element in the (work flow of)
the factory became a key part of an integral grid of cultural, economic and
technological intelligence—such as was built up over time at what became
IBM—by which additional techniques of time control became knowable and
exploitable in other material embodiments and socio-economic spaces.
Similarly, the assemblage of the time control clock exhibits the merged eco-
nomic, technological and regulatory imaginaries of information communica-
tion for management purposes that were concretely developed and propagated
by IBM, leading a once-emerging market in business management now global
in scope.
The foregoing developments in which the time control clock was impli-
cated did not cause or create the explosion in financial practices in the late
20
th
century. They did, however, contribute materially to the conditions for its
possibility. In the 1970s, time control devices worked together with legal and
regulatory interventions to stabilise expectations for a radically new set of mar-
ket practices, as described in the previous section, keyed to temporal specula-
tion and bringing economic ‘rationalization’ to new frontiers. In the 1990s
and 2000s, technical advances in time control and its communication enabled
88 K Knorr Cetina and A Preda, ‘The Temporalization of Financial Markets: From Network to Flow’
(2007) 24 Theory, Culture and Society 116.
89 ibid 117.
90 ibid 123.
London Review of International Law Volume 9, Issue 3, 2021 343
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
the expansion of algorithmic trading. Digital timestamps are not simple repro-
ductions of analogue timestamps, and their development has not been part of
some automatic history or comprehensive plan. Rather, they were produced
by a contingent and peculiar but historically intelligible assemblage of inter-
ests, agencies and things.
CONCLUSION
In conclusion, I want to focus on the conditions by which the infrastructure
for globally standardised time may be favourable (or not) for transformation,
and to consider the efficacy of legal practice at those junctures. There are at
least two material indices in the episode of timestamping that point to possi-
bilities. In that example, mobility and legibility were key. By rendering units of
standardised time legible in mobile form—with numbers and metal wheels,
paper ribbons and coloured stamps, among other things—the time control
clock also produced greater possibilities for translating units of standardised
time into other programs, enrolling time into new programs (such as work-
place control beginning with the factory floor) and reciprocally enrolling new
programs and policy concerns into the infrastructure. Similar material innova-
tions in mobility and legibility apply with respect to the generations of infor-
mation and communications technologies that have replaced the analogue
timestamp, innovations such as are at work in the domain of fintech. The first
section above makes apparent at least one material development resulting
from contemporary innovations in mobility and legibility: multiplication and
distribution of vanishingly brief units of time. The flow architecture of global
finance described by Knorr Cetina was primed by the temporality of the ticker
tape and time control mechanisms, but did not fully come into being until still
more thoroughly temporalised communications, updating at briefer intervals,
became industry standard.
91
Once, the nanosecond was only an abstract possibility; today, it is a con-
crete part of embodied operations of global finance. The latter is more than
the teleological fruition of the former: the embodied unit enjoys a reality dis-
tinct from the abstract one. The embodied unit is also more than a mere incre-
mental addition to or by the technology. Its utilization contributes to new
practices and relationships in time and with time, and in so doing contributes
to a changed temporal order, in which even the valence of pre-existing units
like seconds and minutes changes with the incorporation of the nanosecond
into governance routines. As the temporal order changes, governmental
91 ibid.
344 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
rationalities change with it. As described in the first and second sections, the
relationship is circular: material units of temporal order play a constitutive
role in the production of governmental rationalities, even as governmental
rationalities condition and instrumentalise material units of temporal order.
In part, this is because the temporal order is not autonomous. As the temporal
order interacts with other domains, such as the economic—as its material
units of time are translated into programs operating in these other domains,
and as those domains invest in the production of its material units—it
becomes mutually implicated in the interests and rationalities associated with
them.
Fleur Johns has recently called to ‘reactivate’ competing notions of time
‘as political questions of the first order’ under international law.
92
Her call
points to the need for new ways of recognising and engaging the productive
power and political character of time technologies and temporal regulation.
93
Engaging the productive power of time technologies and temporal regulation
today means intervening in the flow architecture described by Knorr Cetina,
and the in-the-moment register of governance that it promotes. In other
work, addressing international development and humanitarian law, Johns
describes the growing reliance among international institutions generally on ‘a
succession of rapid-fire snapshots resulting from automated dives into vast
and shifting oceans of data’.
94
That reliance enables ‘an open-ended, oppor-
tunistic, now-oriented [governmental] disposition’, such as the flow architec-
ture that the financial world supports.
95
Similar temporalization is evident in
everything from border control installations and institutions of war to food
supply chains.
96
Against this backdrop, Johns has called for a critical practice
to grasp ‘the pattern and formula, storage and transmission’ of ‘the increasing-
ly self-organising streams of digits and “stuff” shaping global affairs’.
97
92 F Johns, ‘The Temporal Rivalries of Human Rights’ (2016) 23 Indiana Journal of Global Legal
Studies 59–60.
93 Cf. F Johns, ‘From Planning to Prototypes: New Ways of Seeing Like a State’ (2019) 82 Modern
Law Review 833.
94 ibid 850.
95 ibid.
96 D van den Meerssche, ‘Virtual Borders: International Law and the Elusive Inequalities of
Algorithmic Association’ forthcoming in European Journal of International Law; M Liljefors, G
Noll, and D Steuer (eds), War and Algorithm (Rowman and Littlefield 2019); J Allen and S Lavau,
‘“Just-in-Time” Disease: Biosecurity, Poultry and Power’ (2015) 8 Journal of Cultural Economy
342.
97 F Johns, ‘On Dead Circuits and Non-Events’ (UNSW Law Research Paper No 19-80, 2019)
<http://dx.doi.org/10.2139/ssrn.3469325>.
London Review of International Law Volume 9, Issue 3, 2021 345
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
How, then, to engage such temporal assemblages that shape global affairs
with self-organising streams of information? In a different register, Nick
Seaver writes that ‘infrastructure is a trap in slow motion’.
98
Seaver posits the
trap in counterintuitive but provocative ways. Drawing on anthropological
studies of traps and trapping, he defines the trap as more than a single-faceted
object to contain a victim. Traps involves persuasion as much as coercion;
they ‘are not just devices of momentary violence, but agents of ‘environ-
mentalization’”.
99
In this latter capacity, infrastructures, likes traps, are ‘mak-
ing worlds for the entities they trap.’
100
The world-making operation is social
as well as material: ‘a trap is not simply the unilateral application of technical
force, but rather a fundamentally uncertain effort to relate to others which
thereby produces a world.’
101
Though not exactly benign—the infrastructure
remains characterised as a trap, after all—neither is it merely sinister. The
world-making character that Seaver observes is productive, and if the infra-
structure trap is not the sort that allows escape, its productive potential might
still be materially recalibrated in order to effect different worlds. In Seaver’s
words, we might ‘move beyond denouncing entrapment and toward reconfi-
guring our captivating social infrastructures.’
102
No trap, Seaver emphasises, is
all encompassing, nor can a trap exist independent of the environment in
which it must entrap its intended captive. Any one infrastructure is itself
caught up in other infrastructures. The elements of the infrastructure for glo-
bally standardised time are enrolled and enfolded in other networks and infra-
structures—fintech, in the example considered here, GPS-driven security
regimes in another—with each assemblage incorporating and contingent on
material parts of another.
Interconnectedness points up the indeterminate and variable condition
of infrastructures, but it also helps to explain their durability.
Interconnectedness in this sense brings with it the design principle that makes
the decentralised network a resilient one. The infrastructure for globally stand-
ardised time is caught up with and stabilised by other traps or infrastructures,
other social and material assemblages, each of them actively stabilised by a var-
iety of other things and assemblages, none of them final or fixed. The factory
98 N Seaver, ‘Captivating Algorithms: Recommender Systems as Traps’ (2019) 24 Journal of Material
Culture 432.
99 Ibid.
100 ibid.
101 ibid.
102 ibid.
346 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
floor has been one material site of interconnection, where economic control of
the mode of production is inserted into the daily rhythms of workers; the
cloud server is another contemporary site.
103
But while the multiplicity of
intersections helps to explain overall durability, they also present sites of mar-
ginal change, where the shifting assemblage of one infrastructure may induce
reciprocal movement in another. Moreover, each interconnection may also
present a tactical site for intervention, where translation exercises from the
one network to another may be contested precisely as the communication
‘switches’ from one systemic orientation to the other. Points of interconnec-
tion and places of translation are also the moments when legal practices typic-
ally play a material role in the assemblage, stabilising the borders of
interpenetrating domains and the interests among them. The mix of atomic
clocks that determine the time standards for global financial transactions are
debated at the ITU-R; while the urgency to stabilise and exploit global posi-
tioning systems, which run on time signals, incentivises the production and
military deployment of miniaturised atomic clocks.
104
In such complex condi-
tions, these constellated social and material points of interconnection and the
legal elements of their interoperation are potential sites for strategic engage-
ment and intervention, whether directed at the infrastructure for globally
standardised time, or at the other networks and infrastructures with which it
is mutually implicated. In addition, fissures among the particular economic
interests implicated in the infrastructure for globally standardised time point
to an instability in the socio-technical hegemony of its operation and repro-
duction. Such fissures present other openings for new interests and (counter-)
engagement in its reproduction.
The constellated points of interconnection between the infrastructure for
globally standardised time and other networks and infrastructures are ripe for
reactivation as political concerns. Time technologies operating near the speed
of light underlie sensationally destabilising developments not limited to high-
speed financial transactions, just as the increasingly material imaginary of
quantum computing points to other opportunities for transformation. The
destabilising character of these developments underscores a moment of possi-
bility—for reactivation, in Johns’ words—in the world-making character of
time technologies at work. But as Kingsbury has stated, ‘the international legal
framework for much of this is at present scanty, woefully lagging, and in
103 Cf. L Amoore, Cloud Ethics: Algorithms and the Attributes of Ourselves and Others (Duke University
Press 2020).
104 DARPA News (n 13).
London Review of International Law Volume 9, Issue 3, 2021 347
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
urgent need of construction.’
105
Economic investments have determined some
of the most powerful technologies incorporated into the infrastructure for glo-
bally standardised time, in the service of Knorr Cetina’s ‘lifeworlds’, but that
does not mean that finance capital enjoys a final or univocal power over the
assemblage. Reactivation, however, is not limited to framework-style regula-
tion of cutting-edge technologies. It may mean opportunistic engagement
with the sort of temporalised flow that the infrastructure for globally standar-
dised time produces.
106
Equally, even at its most technologically advanced
horizons, the infrastructure remains constituted in no small part by everyday
engagement and popular imaginaries. Working up from everyday materialities
can make the possibilities for engaging the infrastructure more apparent by
making them more concrete. Fintech firms buy land to build servers adjacent
to the walls of trading centres, or construct closed transmission systems cap-
able of relaying information at maximum speeds: they take up space in cities,
and are reached or reachable daily by cleaning and construction crews as well
as brokers, bankers, lawyers, mathematicians and security services. Johns
points to the Equinix NY5 data centre in Secaucus, which hosts volumes of
high frequency trading.
107
Data mining operations also set up shop in areas
with cheap energy, cool temperatures and favourable regulatory regimes.
Every one of these spaces includes legal relations and legal practices, among
other things at work in the overall assemblage. And the traps and infrastruc-
tures of which they are a part are constituted out of other traps and infrastruc-
tures, each circumscribed and materially supported by law. In the context of
lethal autonomous weapons systems, also bound up with time technologies,
Sara Kendall has demonstrated how attention to material conditions can point
to old fashioned means of getting ahead of the ‘blink’, or the vanishingly small
moment for intervention in a system predicated on a rapid flow of constantly
changing information, pointing to the power of labour to pre-empt
production.
108
105 Kingsbury (n 3) 181.
106 Cf. A Bastani, Fully Automated Luxury Communism (Verso 2019); N Srnicek and A Williams,
Inventing the Future: Postcapitalism and a World without Work (Verso 2015); A Williams and N
Srinicek, ‘#ACCELERATE MANIFESTO for an Accelerationist Politics’ (Critical Legal Thinking,14
May 2013) <https://criticallegalthinking.com/2013/05/14/accelerate-manifesto-for-an-acceleration
ist-politics/>.
107 Johns (n 97).
108 S Kendall, ‘Law’s Ends: On Algorithmic Warfare and Humanitarian Violence’ in M Liljefors, G
Noll, and D Steuer (eds), War and Algorithm (Rowman and Littlefield 2019) 114. Kendall draws
the ‘blink’ from the work of Brian Massumi: see B Massumi, Ontopower: War, Powers, and the
State of Perception (Duke University Press 2015).
348 Gordon: Engaging an infrastructure of time production with international law
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
A critical step to engage processes of change or transformation, whether
with respect to the infrastructure for globally standardised time itself or some
other implicated assemblage, includes knowing the myriad, mutual implica-
tions of time and law as material indices of intervention, and thereby to recon-
stitute the reality of the infrastructure as a political vector. None of these sites
can simply be reimagined as something entirely other than what they are cur-
rently known to be—their durability militates against this. But they can be re-
engaged and reconfigured in as many ways as there are points of connection
among the social and material agents by which they are known to exist, across
which legal practices play stabilising but contestable roles in myriad material
forms. The constitutive action of the infrastructure for globally standardised
time has been apparent here in its capacity to promote expansive expectations
for lifeworlds in fields of global finance. But I close with the argument that the
corresponding expansion of the infrastructure is also an ambivalent basis for
renewed engagement and intervention in legal practice.
London Review of International Law Volume 9, Issue 3, 2021 349
Downloaded from https://academic.oup.com/lril/article/9/3/319/6529573 by guest on 02 February 2023
