Introduction

London Freedom Seed Bank (LFSB) is a small network dedicated to saving, storing, and sharing plant seeds across London and raising awareness of the political importance of seed saving. It holds a small library of open-pollinated seeds that are grown and harvested by its members and shared at seed swap events or, more recently due to the COVID-19 pandemic, via post. At present the LFSB connects 144 crop varieties, 128 growers, and 99 growing spaces across the wider London area and is coordinated by a small, voluntary steering group who manage, record, and share the seed stock with the wider grower network. For this, the steering group keeps detailed records of the seeds in its care, the people who grow them, and the spaces that they are grown in. In 2020, steering group members Richard Galpin and Anna Clow designed a digital database using the free version of the software Airtable to hold information about the LFSB’s seed stock and to trace how the seeds move through the network through time. Since its inception, this database, called the LFSBase, has grown into a larger and more ambitious project aiming to also record and share the growers’ knowledge and stories alongside the seeds themselves. As Richard explained in a post on the LFSB’s blog introducing the database project: ‘What has become clearer is that the physical seeds are only a part of the resource that we aim to nurture and share at LFSB – there is also an abundance of social relations, grower knowledge, non-human genetic seed knowledge, and political aspiration’.¹

For this chapter, we are drawing on our dual roles as researchers and activist members of LFSB. Our close involvement with the organisation has allowed us to observe and participate in the inner workings of the seed network and, through it, insight into the design process of the LFSBase. Witnessing the excitement with which the LFSB growers and the public engaged with the seeds through the mediating environment of the database, we felt a critical potential in how the LFSBase facilitated both material and digital encounters between the plants and the people who grow them, and how it highlighted the interaction between species and environments that are necessary for things to grow. The data presented in this chapter are informed by both authors’ previous research on the LFSB, but will focus particularly on our analysis of the LFSBase, drawing on interviews and informal conversations we have had within the steering group and with the creators of the Base.

With this chapter we set out to explore how data technologies such as the LFSBase are implicated in the production of knowledges about the living worlds they record. Rejecting the dominant framing of data technologies as epistemically neutral, ahistorical, and merely representational, we follow Eric Nost and Jenny Elaine Goldstein in their articulation of the ways in which data and data infrastructures are material, governed, practised, and fundamentally a ‘product of social, political, cultural and economic circumstance’.² With this understanding, data and data technologies are always already engendered in the particular world-making projects that they arise from; we recognise how data are constantly ‘at the heart of questions of mattering, and consequently at the heart of the political’.³ We want to ‘look beyond data as a “representational resource”, to consider the various forms of epistemic, social and political work that it does and which is done to produce it’.⁴ As recent work and the contributions in this volume extrapolate, processes of datafication of everyday life extend beyond the realm of the human to include the so-called ‘natural’ with increasing intensity. This opens up new ways of reading data itself as a ‘political ecology’, prompting consideration of the ethical implications that this datafication has for how we relate and act in our ecologies and interspecies relations. With this chapter we aim to show how the design and specific practices of capture of information on the more-than-human world is embroiled in epistemological projects, and how digital technologies can both reproduce and subvert the epistemes in which they were created to facilitate other ways of knowing and being in relation to our ecologies.

In the first two sections we examine how practices of recording, ordering, and storing data on the living world are part of a genealogy of imperial knowledge production as part of the establishment of racial capitalism. This has given rise to a seed regime based on ideas of genetic purity and linearity, which is increasingly mediated and governed through data technologies such as national and international databases which commodify the plant life they record as a legible inventory for a global market. Activists resisting this epistemological framing of seeds as intellectual property/commodity avail themselves of database technologies while also subverting the modern corporate seed regime by saving their own seed, often based in an understanding of seed saving and food growing as collaborative, interspecies practice. We then go on to examine how the LFSBase captures the relationships between the seeds in the care of the network, the growers who grow them, and the places they are grown in. We examine how these lively entangled relationships are translated into the digital environment of the database and what kinds of knowledges are facilitated through the design of the database. Understanding how the ‘digitisation of nonhuman life across spaces, species, and scales has profound implications for how we see our place in the world and how we act in it’,⁵ we ask: what kinds of knowledges and relationalities are fostered in the digital space of the database, and what kinds of ethics emerge from these connections? We feel that such a reading of the LFSBase holds generative potential for a political ecology of data which approaches everyday practices of data management as sites of inquiry and knowledge-making where hegemonic epistemes are (re)produced, but also where relations and meanings are made that exist in tension with and leak beyond the strictures of colonial logics.

Technologies of collection

Databases first emerged in the mid-twentieth century from innovations in the office-equipment industry as tools to manage the ever-increasing speed and scale of industrial operations in a globalised market, and soon became the primary technology for organising and storing information.⁶ Yet practices of collecting and ordering data on the living world date back much further. With the expansion of European empires, the disciples of the budding disciplines of natural science were sent to research, record, and map the flora and fauna of the colonies in search for ‘green gold’, scouting and extracting specimens that could prove profitable for empire. This led to an explosion of data flooding into imperial centres, which required new practices of recording, storing, and ordering information, a practice which continues to provide the majority of genetic resources for scientific research in the West to this day.⁷ This led early botanists such as Carl Linnaeus to experiment with ‘different ways of presenting and arranging large amounts of data on plants and animals’ in indexes, tables, and manuscripts in their struggle with the resulting ‘information overload’. They eventually produced taxonomies and nomenclature which became the basis for modern scientific classification today, producing natural history as a ‘prototype of what one could call “data-driven” research’.⁸

These ‘technologies of collection’ were purposefully engendered in the much larger epistemic project of naturalising the extractive operations of capital in the colonies. As plants were violently extracted from their lively local relations they were reframed within the systematised logics of a universal botanical language and rendered ‘legible’ as inventory on a global market.⁹ Local names, relations, and uses of the plants were substituted with the ‘pure’ and supposedly superior classifications of empirical science, erasing all previous modes of being and relating ‘not premised upon the value, profitability and usefulness of plants that underpins the vampiric logic of capitalism towards nature’.¹⁰ Importantly, these same processes of classification not only ordered the ‘plant kingdom’ into families, genera, and species, but they also classified humans through racial taxonomies,¹¹ in order to legitimate colonisation, enslavement, and genocide. The plants extracted from the colonies were cast as discoveries of European explorers and botanists and the inventions of the plant breeders who sought to ‘improve’ them as cash crops for agronomic use.¹² Situating the development of these early ‘technologies of collection’ within the political economy of colonialism facilitates a critical understanding of how data management practices today continue to operate through extractive logics which have material and racialised consequences.

Making modern seeds

This modern dream of ‘improvement’ fundamentally underpinned the development of scientific plant breeding as a cornerstone of industrial agriculture. With innovations in genetics at the turn of the twentieth century, a broader ontological shift has been observed in the life sciences, in which life began to be increasingly understood through doctrines of linear evolution and genetic purity. Lively organisms came to be understood as ‘having an intrinsic genetic identity, sealed off from the vagaries of the environment’, which articulated itself in the scientific search for ‘purified forms of life’, an approach which Christophe Bonneuil and Frederic Thomas have termed ‘pure line ontology’.¹³ This ‘genetic modernism’ reduced seeds to the status of genetic resources and marked a prioritisation of ‘pure’, ‘improved’, and ‘elite’ cultivars that were (re)produced through state-sponsored breeding programmes as governments realised the profitability and biopolitical potential of plant breeding, while farmer-bred and saved seeds were cast as ‘impure’ and unreliable.

This hierarchical ontology was further institutionalised with the development of hybridisation, a breeding practice in which two ‘pure lines’ are crossed to produce hybrid seed, the first generation (F1) of which usually produces a higher yielding and morphologically uniform crop, making F1 hybrids favourable to the demands of industrialised agriculture. However, because F1 hybrids do not produce ‘true to type’ in subsequent generations, they cannot be saved by the farmer for future use. This ruptured traditional farming economies in which farmers would save and exchange seeds from their own crops locally. Instead, farmers were encouraged to buy the more productive hybrid seed from commercial producers each year. Increasing state legislation on ‘variety protection’ further cemented this hegemonic framing of seeds as the intellectual property and ‘inventions’ of breeders.¹⁴ The professionalisation of plant breeding and seed production disrupted traditional farming economies as ‘a new seed regime and a new knowledge production regime’ emerged, in which scientific breeders replaced farmers’ role in crop evolutionary processes.¹⁵

The second half of the twentieth century saw the rise of national and international regulatory frameworks which sought to govern farmers’ ability to reproduce their own seeds by making it practically illegal for farmers to save their own seeds, as commercial crop varieties must be registered on national lists of protected varieties. These registers, mediated largely through databases such as UPOV’s Plant Variety Database or the EU Database of Registered Plant Varieties, have become the primary governance technology on seeds and instruments of globalised industrial agriculture dominated by large agri-chemical giants, three of which control 60% of the global seed trade today.¹⁶ Grounded in the same colonial logics that classified plant life in botanical indexes in the eighteenth century, modern seed registers and their databases determined strict breeding criteria for ‘modern’ varieties, reproducing a view ‘that seeds are the result of individual ingenuity, rather than the collective and intergenerational knowledge that co-evolves with plant genes, soils, and climates’.¹⁷

The streamlining and genetic homogenisation of varieties under industrial agriculture has led to a drastic drop in crop biodiversity and a rise in state-sponsored gene banking endeavours by international research institutions. Through such ‘gene fetishism’,¹⁸ seeds are extracted from their in situ worlds and the multispecies webs of relations in which they usually circulate, along with the different practices, knowledges, and ways of relating that those entail.¹⁹ They are subsequently placed into controlled conditions, based on the assumption that this will allow for more precise understandings of their constitution, behaviour, and genetics, which is vital to industrialised breeding. Managed through database technologies, these gene banks have also begun to include the sequenced genomes of the seeds they hold. This has become a major source of concern and resistance from Indigenous and peasant communities and food sovereignty activists, as it gives rise to biopiracy and continues to reduce the animacy and lively relations of the seeds to sequenced code.²⁰ Parallel to these large-scale ex situ collections of plant genetic material, farmers, growers, and food sovereignty activists have begun politically organising to resist the corporate and state-controlled seed regime by establishing seed libraries and networks for saving and sharing open-pollinated seeds. Some of these initiatives make use of technologies such as databases and digital inventory systems for managing their seed collections and networks.²¹ We now turn to the London Freedom Seed Bank as one such project to interrogate how digital data technologies can facilitate alternative ways of knowing and being in the world to the extractive ‘pure line ontology’ of the modern seed industry.

Knowledge, relationality, and variability in the LFSBase

Before the LFSBase, the LFSB used a simple Excel spreadsheet to take an inventory of the seeds coming in and out of the bank. For Richard Galpin, the Airtable software presented an opportunity not only to record as much information as possible about saved seeds in London, but also to try to understand and visualise the dynamic relationships between seeds, their growers, and growing spaces. In an online interview along with Anna Clow and both authors in February 2021, Richard explained that, with an Excel spreadsheet, ‘there’s no real way of easily seeing how that relates to what happened last year and whether that grower features anywhere else and whether that variety features anywhere else’. The LFSBase offers a means to represent these relationships – and, as we will argue in the following, resists some of the hegemonies of ordering and storing data and seeds that are characteristic of more modernist scientific and commercial approaches. Richard had become familiar with Airtable through his work as coordinator of the Walworth Community Food Hub in London, a mutual aid response to increased food insecurity during the COVID-19 pandemic, and felt it could be useful to update the LFSB’s record-keeping practices. Anna, a computing technician who had come across Airtable through working with the London Renters’ Union, joined the steering group of LFSB in late summer 2020 and worked with Richard (online) to get the LFSBase up and running by November that year.

By outlining how information is recorded in the LFSBase and drawing on Richard and Anna’s reflections on the design and constraints of the technology, we interrogate how the LFSBase is able to hold a multiplicity of knowledges and relations that radically depart from the hegemonic narrative of breeding outlined above. We then go on to examine how variability and adaptability are treated in the database.

Four views: Growers, Growing Spaces, Varieties, and Seed Batches

For Anna and Richard, the principles of accessibility, aesthetics, and ease of use determined their design of the LFSBase. Unlike a standard two-dimensional spreadsheet, Airtable is based around different ‘views’, which allow users to visualise data from several spreadsheets or ‘tables’, depending on the specific information and relationships they want to comprehend. The option of ‘gallery view’ presents the data in tiles, making the software both aesthetically appealing and user-friendly. In our interview, Richard, who is a practising artist, explained: ‘it was [the] gallery [view] aspect that I initially thought, this would be great for seeds, because it just felt like that will be something that would allow us to enjoy the database more and have more of a visual relationship with it’.

The LFSBase records information in four main tables: Seed Batches, Varieties, Growers, and Growing Spaces (see Figure 8.1). Records from these tables are linked, which allows them to be read in relation to one another (see Figure 8.2). As Richard explained in a blog post on the LFSB website in which he introduced the LFSBase: ‘The relationship between these four elements is key. We may have multiple growers in the same growing space, or multiple seed batches from the same grower. A popular variety that starts with one grower, becomes grown by multiple growers. A standard Excel sheet just couldn’t capture this complexity.’

Figure 8.1The four main tables in the LFSBase. (All rights reserved and permission to use the figure must be obtained from the copyright holder.)

Figure 8.2Figure created by Anna Clow to illustrate the linked record fields in the LFSBase. (All rights reserved and permission to use the figure must be obtained from the copyright holder.)

Stories

In addition to capturing technical data about the seeds in the network, the database was created with the aim of facilitating the sharing of practical and experiential knowledge from the growers themselves. Richard continued: ‘The history and the grower’s knowledge are important too. The stories that attach to particular varieties, and the recommendation of how best to grow them, all help to advance the seed’s progression from one grower to the next. Saving the knowledge helps to save the seed, and vice-versa.’²²

When Richard said ‘saving the knowledge helps to save the seed’, he was using an expansive definition of knowledge, which goes beyond the ‘pure’ knowledge of a seed’s genetic profile or its place in a taxonomic hierarchy to express a sense that seeds circulate between places and people and that growers’ embodied and relational knowledge is a crucial factor in saving seeds. Valuing this lay knowledge pushes against the assumptions that commercial breeders have the monopoly on expertise about seeds, and that successful seed saving and breeding should be done in purified ex situ conditions.

Working with Airtable, like any form of data technology, has created dilemmas for Anna and Richard because it involves navigating the assumptions and limitations built into the design of the software alongside their goal of recording diverse information from a range of sources. Richard, at one point in our interview, stated that the Airtable, through its design, forces the user to make decisions, as ‘classifying anything forces you to put it in separate boxes … it forces you to think about, if this, then not that’. For example, when entering a new Variety record, the variety still holds a certain primacy, as new seed batches get linked to broader variety records. However, as we argue below, this also expresses an expansive understanding of what a variety is – Anna and Richard want LFSBase to reflect, and facilitate, the interspecies relations and knowledges at the heart of in situ seed saving, so they are sensitive to the ways in which technologies and their design constraints can work against this. This is another dimension to Richard’s aspiration to share knowledge about the seeds: sharing is a relational act that decentres the individual in favour of the collective.

These challenges are illustrated by the case of the Bloody Marvel variety of lettuce. In the LFSBase, the Bloody Marvel is described under Variety as ‘An experimental new lettuce variety bred by the grower in Walworth, London for resilience to London’s challenging growing conditions. Originally inspired by research into the Bloody Cos variety, also known as Spotted Aleppo originating in Syria in the 18th Century. Blood red splotches on an upright open butterhead-cos’. The unidentified grower in the quote is in fact Richard Galpin, who, as the breeder of this new variety, is a repository of knowledge about it. However, the intimate relationship between Richard and Bloody Marvel is obscured in the Airtable (see Figure 8.3). Anna observed, ‘It [Airtable] does flatten everything, if you [Richard] were to stop growing Bloody Marvel for four years and somebody was to continue, then … from a purely data way, it would almost look like someone else had more of something to do with it’.

Figure 8.3Linked Seed Batches records in the LFSBase Variety record for Bloody Marvel lettuce.

In the days of the Excel spreadsheet, seeds, and information about them, were generally collected on an ad hoc basis, at seed swaps and other events, and Richard described how LFSB co-director Charlotte Dove, who is also a professional gardener, held much of the LFSB’s collective knowledge in her own memory. Richard and Anna are aware of the fact that LFSBase is just one element of the ways in which people interact with seeds and each other, that databases can be alienating and spending time in front of a computer inputting data is not necessarily what growers enjoy about seed saving. Richard observed that, to tell stories about the seeds in care of the network, ‘it needs both narrative and, probably, interaction. These stories need an audience, right?’ So, while he is very aware that, as he put it, Airtable is ‘not going to write the stories for us, but it’ll give us a box to put them in’, he is also pragmatic about the fact that this is not its primary purpose, as is reflected in its design. Yet, when we probed him a little more on this issue, he did observe that, since the LFSBase had been established, most seed orders were for those varieties that have ‘more background story in the description’. He suggested that this is linked to the fact that people do not rely on the LFSB for their entire seedstock, so are specifically ‘looking for something unusual or special that they wouldn’t get somewhere else’, so here the description or story of the seed, contained within its particular ‘box’ in LFSBase, becomes a determining factor in which variety they decide to grow.

The LFSB’s network values ‘stories’ about seeds as knowledge, but as Richard and Anna reflected, this is limited by the fact that Airtable is not designed to capture stories. Airtable is based around having ‘boxes’ to put things in, as Richard described it, and with this go certain assumptions about the absolute nature of categories. By contrast, the design of the LFSBase aims to facilitate, rather than just capture, relationality. The database is envisioned as a mediator of multispecies relationships, driven by the key ethic of sharing upon which the LFSB is built. This decentring of individual authorship and ownership again contrasts with the ontology of modern seeds as deployed by industrial agriculture.

Diversity

As ‘the biggest and most complex field’ in the database, as Richard described it in our interview, Seed Batches is the category upon which most of the operations of the database are built. Every new donation of seeds to the seed bank is registered as a new Seed Batch with a unique batch code, which allows the record to be traced through the database and through time. This record is linked to the Variety that it falls under, but captures more specific information about the particular batch at hand, such as how much light or water the crops received, whether the growers selected the healthiest, most vigorous plants for saving, and whether the plants were isolated to prevent cross-pollination – a necessary step for most crop families to ensure the varieties saved reproduce stably over the generations. In addition, growers can comment on the growing process and add memories or reflections in the ‘batch notes’. Seed Batches are in turn linked to the records of the grower who grew and donated the seeds and the growing space where they were grown. As such the Seed Batch records emerge from the convergence of the seeds, the growing space, and the care of the grower not only physically, but epistemically within the LFSBase.

Varieties are made up of multiple Seed Batches, which in turn each represent a unique assemblage of relationships from which they have emerged. With each new Seed Batch, the Variety record expands to encompass new information about that variety. As Richard said in our interview, ‘What we said about that variety two years ago is probably pretty similar to what they would say about it two years later. But any new information they create becomes part of the record of that variety for us. We subsume that information into the totality of information we have about that variety’. In this, the LFSBase fundamentally departs from the fixist paradigm of the modern seed industry by which varieties have to be ‘distinct, uniform, and stable’ to be considered agriculturally ‘useful’ crops. Rather, in the LFSBase, Varieties act as a container for the multiplicity of Seed Batches, the knowledge related to them, and their relationships to people and places.

In his blog post, Richard explains this distinction as it figures in the LFSBase:

The example of amaranth illustrates this dynamic use of varieties further. In contrast to the relatively controlled knowledge about Bloody Marvel, amaranth has a fluid relationship with the LFSBase’s data ‘boxes’. Amaranth, a crop that is grown around the world and known for its incredible resilience, is grown by many Londoners, though it is generally less familiar to white growers. As they were compiling the LFSBase, Richard and Anna realised that amaranth had several different identities in the LFSB’s records, having also been recorded as different varieties with names including ‘latte’, ‘callaloo’, ‘amaranthus’, and ‘tricolor’.

While some seeds in the LFSBase are accompanied by much grower knowledge, the proliferation of amaranths and uncertainty about whether they were different varieties instead demonstrates a lack of knowledge about this particular plant among the (majority white) steering group and growers’ network, for whom it is a largely unfamiliar crop. Richard noted, ‘Varieties aren’t really used in the same way, it seems, within callaloo so much, culturally’. So, while the LFSBase is attempting to take a diverse approach to data curation that is inclusive of non-expert knowledge, when there is a lack of knowledge about a plant, what scant information there is can spill out of its boxes and potentially become confusing or misleading. Richard, in consultation with others in the steering group, decided to split up the seeds saved from the original seeds by their main features, resulting in five Variety entries. This has led the LFSB to conduct a series of ‘amaranth trials’ with a number of growers who agreed to grow out some of the seed to try and establish the different characteristics within the varieties. The information gleaned from this experiment will then be fed back into the Airtable in order to improve the knowledge about amaranth in the database.

Seeds in the bank are expected to have some genetic variability, partly out of pragmatism about the ‘uncontrolled’ conditions of home and allotment growing and partly out of a sense that open-pollinated seeds should be allowed to interact with their environments. As well as decentring individual growers and expert scientific knowledge, the LFSBase draws attention to the complex and interwoven relationalities involved in saving and sharing seeds, and unsettles the pure line ontology that epistemically produces seeds as distinct and innate genetic packages. This is further illustrated by the Mamadoli Pumpkin variety of tomato held by the seed bank:

The brief description of the seed variety is clearly not one that is bounded by the rigours of genetic purity, or indeed a strict interpretation of variety, but instead reflects a multispecies entanglement between people, place, and plants. While more formalised seed banks might reject this seed because of its uncertain identity, in the LFSB, it is accepted and given a name which reflects its diasporic heritage, memorialising the connection to the original donor’s family and country of origin. Here, the fact that the variety ‘didn’t have a name’ is recorded as part of the knowledge held about this seed.

When, in contrast to standard commercial practice, seeds are open-pollinated, they are inherently dynamic and adaptive. Each generation receives genetic information from their parents, making them able to adapt to changing environments and thus ecologically resilient.²⁴ This genetic diversity, and the ability to reproduce true-to-type in subsequent generations, is what distinguishes open-pollinated seeds from F1 hybrids. Because the seeds of the LFSB are grown in situ, they are able to adapt to local growing conditions with each generation, preserving the agro-biodiversity unique to London ecologies. This genetic diversity of OP seeds is reflected in the database through linked record fields in which the Varieties are linked to the records of the Seed Batches that make up that variety stock.

In this practice, the growers, and their corresponding ‘Grower’ records in the database are not given primacy over the other elements in the database, reworking the modern narrative of agriculture as a manipulation of ‘nature’ by the human ‘grower’ protagonist/subject, giving way instead to an understanding of cultivation through its constituent interspecies processes.

In our interview, this tension was framed by Richard as a question of scale, or ‘granularity’ as he put it. Because open-pollinated seeds are able to adapt to their environments with each generation, if a particular variety is grown out over multiple generations by the same grower and in the same space, it would adapt to that particular ecology. This raised the question of to what degree the ‘London-adapted varieties’ in the care of the LFSB would be specified. Richard posed the question: ‘We often talk about London-adapted varieties, but by the same logic, do you take garden-adapted varieties? So you say well, okay, this seed batch was growing in this garden in these conditions, therefore do we want to mix it with the other version? … do we consider it the same or do we now consider it different?’ This illustrates the fundamental tension between genetic purity and diversity that many seed savers face. It also draws attention to the fragility of the epistemic construct of varieties. With these questions of scalability, the LFSB also negotiates the market logics of specification and homogenisation presented by F1 hybrids, which are bred to reflect the epistemic construct of botanical varieties or cultivars, understood as distinct, uniform, and stable. Any uncharacteristic traits for that particular variety are ‘rogued out’ and, as such, a pure-line ideology of varieties is produced. However, this roguing out is also an important practice of ensuring that the particular traits a variety has been bred for over generations continue to be passed on, which is why seed saving requires both roguing out and for most species isolating varieties from each other to prevent cross-pollination and thus a ‘contamination’ of the seed stock. At the same time, maintaining genetic diversity within a variety is important, as lack of diversity can lead to ‘inbreeding depression’, that is, a generational decline in health and vigour of a variety. After all, it is the genetic diversity within varieties that makes open-pollinated seeds more resilient and adaptive. This fundamental tension between genetic purity and genetic diversity is often raised in seed-saving circles, and approaches, opinions, and rigidity vary.

While this tension may be debated at length in grower and seed-saving circles, many also find that, more often than not, seeds behave differently from a grower’s expectations. Richard articulated his approach as one of ‘happy accidents’, and indeed, the LFSB’s stock includes varieties that have come about through unforeseen or unintended crossings and evolutions, such as the lettuce variety ‘Pandemic’, which has the following description in the LFSBase:

Furthermore, what to save and whether to isolate often comes down to the individual grower and to subjective or experiential factors, such as taste, beauty, practicality, and so on. As such, seed saving in the LFSBase is understood as experiential, subjective, and dynamic and, while the basic aim is to maintain the varieties in its care, the LFSB happily includes new, different, or unusual seeds in its collection, demonstrating an approach to genetic diversity akin to decolonial approaches to human–plant co-evolution.²⁵

This focus on representing the multispecies relationships that make up the LFSB network sets the LFSBase apart from imperial and commercial uses of recording technologies. The seeds in the network are represented through their constitutive relationships, as the Seed Batches unite the different elements that make it possible for seeds to grow and be shared. From this, an understanding of seed saving and cultivation emerges as a conjunction of ‘intra-species skills that are cultivated in a dialogue between plants and humans that also involve … microorganisms, soil, weather, geological conditions and so on’.²⁶ The seeds, as all ecological beings, ‘do not precede their relations with others, rather they constantly emerge through material relations’.²⁷

Conclusion

The LFSBase tells an interesting story about the tension between how technologies are imagined to function by their creators and how they are put to use by users. As this shows, while databases are classificatory technologies that bend towards order and control, they are also being used in unanticipated and subversive ways to effect more radical ends. In one sense, then, the LFSBase works the system, using a software program to facilitate activities that push against political–economic orthodoxies in favour of building multispecies communities, sharing knowledge and experience freely, resisting corporate control of biotic materials, and questioning the need for stringent controls over genetic purity.

The LFSBase offers insights into the epistemic and material implications of digital technologies on human–plant relations. While technologies of collection have historically been employed for imperial purposes and continue to buttress a seed industry based in the colonial disavowal of the relationality and lively dynamism of cultivation practices, the LFSBase demonstrates how such imperial tools can be used by activists resisting these systems on the ground for generative and liberatory aims, facilitating alternative ways of knowing and being in relation to the more-than-human and our species’ place in agricultural practices. Resisting the genetic modernism of the industrial seed regime, the LSFBase facilitates an understanding of seed saving and growing as a relational collaborative interspecies practice between people, places, and plants.

Our reading of the LFSBase intends to add to a discourse on how increasing datafication impacts our relationships to our ecologies. Public discourse rightly identifies the challenges that come with these developments, but the LFSBase demonstrates how digital technologies can play an active role in the creation of alternative meanings and relationships beyond the knowledge systems they were built in. Data software can be a practical and integral tool for facilitating the material exchange of seeds and the gathering of stories and knowledges that might normatively be discounted. A range of activist projects such as the Ida B. Wells Justice Data Lab share such an approach and already make use of data technologies for liberatory ends. The LFSBase offers an example of how grassroots organisations can use data technologies to ‘design anew’ the worlds that we want to live in by offering infrastructures for imagining alternatives and acting as tools for living them. Repurposing the kinds of imperial tools that have historically been used to institute classificatory hierarchies that have divorced us from the interspecies processes of which we are a part, the LFSBase disrupts the prevailing market logics of racial capitalism such as individualism, ownership, and improvement in the modern seed industry and promotes a relational ethic of seeds, growers, and ecologies as co-constitutive, coevolutionary. It thereby gestures towards a radical otherwise.