Why Reinvent the Wheel? Materializing multiplicity to resist reification in alternative organizations

Identifying possibility with achievability

The achievability faction is led by the Paris-based Project A, the first and largest of the RAL projects. Project A invests great effort in materializing an alternative grocery, demonstrating its dereification of the dominant model. It is thus clearly committed to the real possibility of a model that has not previously been realized. Nevertheless, key members of Project A are quite unambiguous in their estimation of the material constraints on this possibility: ‘Anybody can say that they’re doing these big philosophical things, but are you really doing it? [. . .] Ideas are second, reality is first’ (A1). In this way, Project A’s pursuit of an alternative sociotechnical configuration is shaped by scepticism regarding mere ideas of what is possible.

To ground its pursuit in material reality, therefore, Project A carefully studies PSFC, which it views as the best existing materialization of a solidarity grocery. The achievability of PSFC’s model is demonstrated by its endurance: ‘We think we’ll be stronger if we follow a model that works and don’t try to reinvent the wheel. [. . .] 43 years on, it’s the only one that has lasted like this, it’s the only one that works’ (A2). Project A attributes this unusual persistence to PSFC’s mastery of the material constraints bearing on alternative organizations, developing a model ‘built on pragmatism and professionalism’ that is ‘efficient economically and politically’ (A4). On this account, the PSFC model is successful because it is designed to function effectively within its context, including the relevant economic and political environment. Project A uses the PSFC model as the basis for its own Paris model on the understanding that many of the same material constraints will apply: ‘There are a lot of things, in fact, that are quite logically established, and if we had to do it again [. . .] we would come to the same conclusions as Park Slope’ (A2). Project A thus identifies the set of possible alternatives with the set of those achievable according to existing constraints. Furthermore, by suggesting that they would likely independently reproduce the PSFC model through logical analysis, Project A implies that the set of achievable alternatives is quite small.

Identifying existing constraints on alternatives

Of course the French context is not identical to that of PSFC, so the achievability faction must identify which constraints apply. This assessment is shaped by the identification of possibility with achievability. Regarding social constraints, for instance, in the context of French working adults’ limited leisure time Project A posits that there is a ceiling on how much energy participants can invest in a solidarity-based project: ‘You could be a really good person and naively say, “I think that members should have more [decision-making power]” – not realizing that probably most members [. . .] don’t want to decide on every little thing’ (A1). Project A also appeals to common sense regarding consumer behaviour, suggesting that the grocery must offer a large range of products ‘because only a few activists are going to flagellate themselves long term to have to work for the right to buy pasta and black radish’ (A4). Furthermore, given risks of incompetence or abuse, Project A identifies constraints regarding the trust placed in ordinary participants. Arguing that a solidarity grocery cannot operate without a minimum number of employees, for instance, Project A cites purchasing stock as ‘a real job’ and ‘a complex task to delegate to members’ (A4).

In terms of institutional constraints, the achievability faction is highly cognizant of the laws to which a solidarity grocery is (potentially) subject:

From a legal point of view we must be able to prove that we only sell to our members. This is a vital point for cooperatives since the fact that we have the right to use volunteers in a commercial structure is only tolerated on the grounds that we are not in competition with other shops. (C4)

Proposed alternatives are thus evaluated in terms of their coherence with these legal constraints versus their vulnerability to sanction. Nevertheless, as a novel sociotechnical configuration, the Paris model is intended to influence institutional constraints going forward. Participants aligned with the achievability faction thus draw on theories of institutional change, highlighting in particular the potential for the RAL to wield its own institutional power as a formal federation that ‘would be able to promote the model’ (D2). There appears to be consensus among this faction that such a federation would require agreement on ‘a common course’ because ‘there is strength in unity, not division’ (J1).

Regarding economic constraints, the achievability faction emphasizes the competitive nature of the food and grocery sector, noting that, while a solidarity grocery is alternative, ‘it also remains a company with the same problems as the others, [. . .] constrained by a certain number of things. [. . .] The context in which we operate – globalization, excessive competition, price wars – we won’t escape that either’ (E2). Adherents of this approach thus emphasize the importance of securing operational efficiency: ‘You have to be pragmatic. Do the fundamental work. Ensure the sustainability and economic viability of the supermarket’ (A1).

Finally, in terms of technological constraints, Project A emphasizes the barrier to alternatives posed by the strong relationship between mainstream ERP software and the dominant grocery model, expressing deep concern that it would ‘have to adapt’ its operations to suit such ‘software that you’re going to find on the shelf’ (A2). Yet an advantageous feature of the technological context is the availability of open-source software, such as OpenERP, that enables projects ‘to modify the ERP to meet the specific needs of our model’ (A3). Constraints linked to ERP software’s alignment with mainstream groceries are thus attenuated. On the other hand, the modification of ERP software comes with ‘a regulatory constraint’ (H2) in the French context due to a 2018 law requiring ERP software to be certified to prevent customizations that would enable fraud. Such certification must be renewed each time the software is modified, imposing significant costs on revision. As we will see below, the achievability faction does not view this limitation as a reason to reject ERP software, in part due to its perception of a broader technological constraint: the tendency for technology to converge on a single most efficient solution. Some adherents of the achievability faction articulate a softer version of this technological determinism, suggesting ‘we’re going to ask ourselves exactly the same questions, and we’ll often find the same solutions, so we might as well save ourselves some work’ (G1), while others put the point more sharply:

Why reinvent the wheel by shifting the costs of implementation [. . .] to your co-ops rather than using what already exists? [. . .] I have the impression [. . .] that some people within the cooperatives want to please themselves by tinkering with their own solution to the detriment of the general interest of their cooperative. (A4)

Here the suggestion that it is pointless to explore alternative technological solutions is accompanied by a moral critique, positioning ‘reinventing the wheel’ as self-serving behaviour. This is a key theme in the achievability faction’s efforts to materialize its particular alternative, addressed in the next section.

Materializing a particular alternative within existing constraints

Project A constructs the Paris model in relation to constraints it deems necessary to maximize the materialization of the solidarity grocery. Most obviously, the Paris model imposes social constraints within the grocery through strict rules to guard against the above-identified risks of incompetence or abuse of responsibility. Many of these are imported directly from the PSFC model, including the hierarchical management of the store by employees and the imposition of sanctions on members who miss their work shifts. As seen above, however, Project A also attempts to impose social constraints across the RAL by invoking a moral obligation for projects to respect material constraints: ‘You have a responsibility to your members to base what you do on experience. [. . .] If you base it on just what comes out of your heads, that’s lazy and irresponsible’ (A1). According to the achievability faction, this entails a duty to develop a single shared account of the optimal solidarity grocery model:

We have a responsibility to the people who meet us and they say, ‘how do you do this?’ [. . .] We don’t want to say, ‘you know, but there’s a dissenting view from this group and another dissenting view from this group’. (A1)

One aspect of this shared account is a certain interpretation of the institutional constraints bearing on solidarity groceries, such that it is considered prudent to incorporate as a cooperative rather than remaining an ‘association’ (non-profit club). These norms are then reproduced by other adherents of the achievability faction: ‘We try to say to everyone we help that it’s important to respect the conditions [. . .] [Project A] knows much more than us about this because they created the legal sketch’ (C2). The Paris model further recommends establishing material relations with institutional bodies, particularly French and European institutional funders, which serve to materialize the solidarity grocery while imposing certain constraints on the development paths it can then pursue. Project A claims that where projects fail to attain institutional funders’ support ‘it means that the project is not viable’ (A1) – eliding the viability–achievability distinction and thereby demonstrating the identification of possibility with achievability.

The various material constraints on achievability are often consolidated around economic performance specifically: ‘This may sound simplistic but the main indicator is turnover. [. . .] Everything else is secondary [and] must be set aside until the viability of your supermarket is secured’ (A4). Indeed the achievability faction positions economic success as a precondition for dereifying the dominant grocery model and demonstrating ‘the reality that cooperating is more powerful’ (A1). The Paris model is thus designed to establish economic resilience through high operational efficiency, taking from the PSFC model an emphasis on economies of scale and therefore proposing that projects must attain a specific minimum store size to cohere with economic constraints. This centring of scale and economic performance as a baseline for achievability is reflected in Project A’s assertion that only projects successfully operating a supermarket have legitimacy to shape the Paris model:

This idea that people really quickly wanted to start a federation [. . .] where we all decide on the direction that we’re going to go – for us, it was psychotic [. . .] These people are in no position to decide what we do as a group of coops because they don’t have a supermarket. (A1)

Finally, we can understand AltERP as a means of attenuating, via sociotechnical revisability, the technological constraints of mainstream ERP software. AltERP is an open-source software built on the OpenERP framework and custom-designed by Project A to facilitate the operational efficiency of the Paris model. The modularity of OpenERP means that irrelevant ERP elements (e.g. modules related to marketing promotions) can be discarded and replaced with custom modules unique to the solidarity grocery model (e.g. modules to manage members’ work shifts). At the same time, AltERP materializes an alternative set of technological constraints due to the integration of these modules into one unified system designed to secure the uniform realization of the Paris model. In this way, AltERP materializes a meta-design approach to the solidarity groceries’ sociotechnical configurations, specifying a particular standardized architecture informed by constraints on achievability identified and interpreted by Project A’s development team. The Paris model’s organizational ‘rules are implemented in [AltERP]’ (A1) including, for instance, the ‘complex rules of [. . .] status and right to make purchases’ (A4). AltERP ensures compliance with institutional constraints by, for instance, connecting the checkout to the member management module to ensure only active members may complete purchases. Furthermore, AltERP materializes economic efficiencies through the integration and automation of various store management functions. By writing these constraints into the software, Project A already specifies many of the relations that will constitute any grocery adopting AltERP.

As an open-source software, AltERP formally permits project-specific customization within the bounds of this broader architecture and can thereby contribute to the materialization of multiple distinct organizational models: ‘The idea is that we have something that’s a base program [. . .] the essential things that we need to get done so that it works functionally for everybody’ (A1). Yet Project A generally cautions against such customizations, emphasizing the advantages of a single shared tool: ‘There are some who [. . .] think that everyone, given that it’s open-source software, can do what they want with it [. . .] and as a result, [AltERP] loses a little of its possible strength’ (A2). For projects that similarly identify possibility with achievability, adopting AltERP as a means of reliably implementing the Paris model is a sensible choice because ‘we’re not going to reinvent a system’ (I1) and ‘we benefit from [Project A’s] feedback and we can thus move forward more quickly’ (F1). Indeed Project A strongly encourages the adoption of AltERP by the other RAL projects, offering this software for free as a means of growing the relational network materializing the Paris model: ‘We can now offer the tool to other cooperatives to help them structure themselves’ (A4). In this way, the achievability faction tends towards concretization of the Paris model, particularly via the AltERP software.

Identifying possibility with viability

The viability faction explicitly rejects suggestions that the Paris model is the only viable alternative: ‘I understand that it is difficult to admit it, but other models of cooperative supermarkets are possible’ (B5). This faction objects to the inhibition of exploration of further possible alternatives beyond the Paris model: ‘the proposal is very clear-cut, it doesn’t leave much room for debates’ (H2). These projects also express doubt regarding the achievability faction’s analysis of existing constraints: ‘we don’t think that [the Paris model rules] guarantee you can succeed: some of them are really important, sure; some of them are not really important’ (B2). This scepticism is partly rooted in questions regarding the stability of existing constraints. For instance, some view their solidarity groceries as part of a broader sustainability movement ‘transitioning towards [a society that is] more social, less carbon-intensive, etc.’ (D1). These projects thus recognize a diversity of social orders against which the viability of a political project can be assessed, and suggest that it may be more prudent to align with future societies’ constraints. In this way, the viability faction highlights the fundamentally political nature of interpreting constraints on possibility, and this awareness underpins their particular sensitivity to the reification of existing constraints in the AltERP software, noting that ‘tools are always easier to share than political aims, because a tool is just a tool – and that’s why it can be dangerous, because the tool can hide divergences on other things’ (E2). They thus propose an alternative assessment of the constraints on solidarity groceries, focusing on existing constraints’ contingency and space for experimentation, as well as novel emerging constraints.

Identifying contingent constraints on alternatives

Latitude regarding the social constraints identified by the achievability faction is demonstrated by, for instance, Project B’s experience of flouting the Paris model’s rules: ‘We were told “cash, don’t do it”, well, we do cash. We were told “there always has to be an employee who controls the cash”, well, it’s the volunteers who close the cash register’ (B3). Other projects take this evidence into account when assessing their grocery management options: ‘[Project A] promotes the idea that it’s impossible to start this project without paid people [. . . but Project B] prove that it’s possible to make the project with only one’ (D1). To explain this apparent shortcomings of the achievability faction’s analysis, viability faction adherents highlight the limited applicability of consumer behaviour theories since ‘our members do not join for “a competitive service”. They join to co-construct an alternative model that reflects them’ (B5). The social constraints written into the Paris model thus reflect a particularly narrow conception of the relationship between the grocery and its members.

The viability faction further questions the institutional constraints identified by the achievability faction, noting that ‘if you don’t want something to be done, you can always invoke a rule, a regulation, something, to say that it shouldn’t be done’ (E2). For instance, these projects debate the degree to which the law regarding the commercial use of voluntary labour actually applies to solidarity groceries: ‘there is no really precise law about [whether] you can be a volunteer in your own coop’ (C2). Furthermore, some institutional constraints only apply where a project ‘exceeds a certain surface area’ (H1), and are thus contingent on the Paris model’s minimum scale requirement.

Indeed, the viability faction disputes the idea that this scale requirement is determined by economic constraints: ‘Often [Project A] repeat that “it’s the only way possible”. [. . . B]ut we make profit. So like we can do it’ (B1). They further highlight the apparent changing relationship between scale and economic success in the food and grocery sector: ‘Nowadays [. . .] the hypermarkets are losing money. [. . .] They are quitting these huge markets and they’re moving into towns, into the neighbourhoods [with] small shops. So why don’t we do that as well?’ (B1). On this account, it is the Paris model’s scale mandate that generates demanding economic constraints via loans, expensive rental contracts, and so on.

The contingency of economies of scale also informs the viability faction’s assessment of the technological constraints associated with OpenERP. These constraints include the high cost of OpenERP development due to the 2018 software certification law and the fact that ‘competent [OpenERP] developers are scarce and expensive’ (C4). Additionally, the complexity of ERP software constrains the degree of democratic decision-making possible as ‘it can only be mastered by experts, [. . .] and really only a few can understand what it can do and how to use it’ (D1). The parameters of genuinely technological constraints are thus visible only to those engaged with the software development, and therefore while AltERP ‘is offered freely, [. . .] it seems to me like a trap that’s going to close on anyone who sets foot in it’ (D1) due to its developers’ firm alignment with the Paris model. Given that these ‘tools are often “obstacles to doing”’ (E2), the path dependency of sociotechnical configurations involving ERP software means that such contingent technological constraints can nevertheless block projects’ ability to explore further alternatives.

Materializing multiple alternatives beyond contingent constraints

The viability faction thus pursues alternative models that take advantage of the contingency of existing constraints and experiments with alternative configurations of constraints aligned with their own political commitments including, as we will see, a commitment to sociotechnical revisability. Viability Faction projects invoke, for instance, social constraints rooted in democratic norms as reasons to reject the hierarchy of the Paris model, emphasizing that ‘participation is at the heart of the project, it is the very reason for its existence’ (B5). According to these ‘self-organization ideals’ (D1), solidarity groceries must foster social relations that position members as participants rather than merely consumers:

You can’t say that the shop belongs to everyone if you don’t trust everyone to run the checkout. In fact, you have to take it the other way round, I always say ‘You have to get people to think that the supermarket belongs to all of them and that when they steal from the till, they steal from themselves.’ (B3)

In so doing, the viability faction works to transform the social constraints identified by the achievability faction.

This faction similarly aims to transform existing institutional constraints by, for instance, adopting organizational forms aligned with the spirit of the law, if not the letter:

We don’t have the right to operate illegally, of course we don’t, but at the same time, when you’re involved in projects that move the lines a bit, that want to transform things – I mean, at some point it’s the ‘why’ that’s most important. (E2)

One way to facilitate this form of experimentation is to avoid configurations that would trigger legal constraints, or their enforcement. For instance, many viability faction projects opt to register as ‘associations’ in part to avoid the constraints of the cooperative status ‘because we are permanently evolving, so we don’t want everything to be written into marble’ (C1). Against achievability faction claims regarding the institutional power of federating, adherents of the viability faction highlight that ‘if one coop has a legal problem [. . .] they can attack all of us coops if we say that we are the same’ (C2). By contrast, therefore, project independence can support experimentation beyond existing institutional constraints by mitigating the RAL’s vulnerability to legal challenge.

It is because many viability faction projects do not aim to become high turnover businesses that they can resist the scale that would trigger institutional constraints. Such voluntary economic constraint is termed ‘financial sobriety’ within this faction, and is promoted as a means of minimizing financial obligations so as to facilitate exploration of viable alternatives: ‘Our project, however modest, has been viable since the grocery shop opened in 2017. [. . .] As this objective of economic viability has already been achieved, the project focuses on co-construction [. . . of] an alternative model’ (B5). Financial sobriety is a principle shared with other solidarity economy projects, such as local currencies, and the viability faction posits that fostering relationships with such alternatives can mitigate the constraints imposed by market competition, instead contributing to the materialization of a solidarity economy.

Finally, against the Paris model’s single shared technological solution, this faction develops multiple solutions in parallel so as to explore a fuller range of viable sociotechnical configurations: ‘We have different solutions so the new supermarkets can choose which one is best for their needs [. . . T]hat means that there is some vitality, there is room for discussion, choice, etc.’ (K1). To facilitate such choice, the viability faction places heavy emphasis on sociotechnical revisability as a means of attenuating the constraints of both technologies designed for mainstream groceries and those designed for solidarity groceries, particularly AltERP. One set of viability faction projects adopts a meta-design approach that affords more revisability than does AltERP. Of this set, some projects use CustomERP-SaaS, a software-as-a-service solution built on OpenERP by a third-party developer offering tailored module customizations, enabling projects to experiment with different configurations more cheaply and easily. Where projects want to follow different accounting practices, for instance, tailored accounting modules can be developed. Other projects develop software to interact with the ERP via an application programming interface (API), which can thus be written ‘in more common programming languages, the developers of which are more accessible and less expensive than [OpenERP] developers’ (C4). This reduces barriers to experimentation regarding the stores’ overall sociotechnical configurations. These viability faction projects thus materialize distinct sociotechnical configurations aligned with a meta-design approach to sociotechnical revisability.

Yet this meta-design approach renders certain technological constraints inaccessible for revision and thereby vulnerable to reification. For example, modifications of ERP systems are subject to the 2018 software certification law, imposing constraints on sociotechnical revisability. For this reason a second set of viability faction projects adopts an atomistic design approach, using stand-alone tools to fulfil distinct store management functions in place of any integrated ERP-based configuration. Such atomism reduces technological constraints on revision as it entails ‘no connection to the rest (voluntarily) [permitting] a lot of flexibility (new developments are quick and free)’ (B5). For instance, Project B is able to rapidly prototype and implement a configuration that reduces the project’s reliance on employees by enabling ordinary project members to open and close the store with a passcode that is updated daily, using a cheap lockbox and a simple script on the member management platform. Atomistic design thus supports sociotechnical revisability via both the reconfiguration of relations between atomistic elements and the modification of each element individually.

Nevertheless, some viability faction adherents note that even the modification of atomistic elements – like the script on the member management platform – risks violating democratic norms by allowing power to accrue to those with technological expertise. These more computer-literate members may present their opinions regarding technological constraints and possibilities as relating more ‘to technology than to politics – that, I think, is dangerous’ (E2). For this reason, many viability faction projects adopt an atomistic approach incorporating off-the-shelf proprietary technologies: ‘we went with [SimpleCheckout] that costs 50 euros/month + purchase of equipment (5000 euros) and that’s it!’ (D1). Particularly where participants describe choosing ‘an existing solution [. . .] so as not to reinvent the wheel’ (B5), this account may be seen as a reversion to reification of mainstream sociotechnical configurations against the sociotechnical revisability of projects adopting OpenERP solutions in both the achievability faction and viability faction. Yet while the proprietary elements are not themselves revisable, these projects emphasize the reconfigurability of relations between atomistic elements in that they can still ‘export our data and rent something else or buy something else’ (D1). In this way, these viability faction projects materialize multiple additional sociotechnical configurations aligned with their democratic norms and interpretations of the constraints of sociotechnical revisability.

Mirroring the limitations of the meta-design approach, the atomistic design approach materializes a significant technological constraint by prohibiting the integrated sociotechnical configurations needed to realize more complex functions and large-scale operations. That is to say, both approaches demonstrate distinctive limits regarding sociotechnical revisability. We would thus expect each approach, when pursued in isolation, to ultimately reify those elements of the sociotechnical configuration inaccessible to their mechanisms of revisability. Yet within the RAL these approaches are not pursued in isolation: they are materialized in parallel by highly communicatively connected projects. The viability faction, in particular, notes that ‘it is very positive that projects that feel the need are working on alternative solutions’ (C4) and so establishes shared resources to encourage information-sharing and experimentation. By pursuing multiple viable sociotechnical configurations simultaneously, the RAL inhibits the reification of any one solution. We call this strategy for resisting reification materializing multiplicity.