Cultural capital and public libraries in the era of algorithmic collection management

Cultural practices have traditionally been understood as tightly linked to social class. Largely in the wake of Pierre Bourdieu's (1984) theories and empirical studies, social class is seen as the ultimate shaper of cultural practices. Bourdieu postulated that the upper classes possess a “legitimate” taste – that is, a like for highbrow culture, such as classical music or poetry – while the middle classes emulate the upper classes without ever properly succeeding in doing so, and the working classes have a popular taste – that is, a preference for things shunned by the upper classes, such as schlagers or romantic novels. While Bourdieu's view has received much criticism, recent scholarship has broadly confirmed that cultural practices are organised according to social hierarchies in almost all of the Global North. That is, highly educated, resource-rich groups participate in culture more actively and broadly than groups with fewer resources (Bennett et al., 2009; Katz-Gerro and Jaeger, 2013; Purhonen et al., 2019; Reeves and de Vries, 2019).

The novel challenge for Bourdieu's rather deterministic idea is derived from the rapid increase in machine-based cultural recommendation systems. It has been speculated that during the last decade or so, algorithms have shaped our cultural practices in new, radical and unforeseen ways and are transforming the traditional top-down models of criticism and recommendation (Beer, 2017; Cohn, 2019; Hallinan and Striphas, 2016; Striphas, 2015). At the same time, there has been widespread discussion on the challenges that the ubiquity of algorithms poses for democratic culture. It has been argued that while algorithms represent certain machine-based neutrality and “crowd wisdom” (Striphas, 2015) loaded with positive connotations of democracy, participation, belonging and so on, the normalisation of algorithms could actually backfire through the abandonment of public or democratic culture and the emergence of polarised and narrowed-down cultural repertoires.

I argue that public libraries play a key role in this broad transformation of cultural practices as they start embracing AI-based recommendation systems in their collection management (Cox and Mazumdar, 2022; Cox et al., 2019). This stands out as particularly interesting and potentially contradictory because public libraries are central actors in levelling cultural and social inequalities. They provide safe and equal spaces for the public, regardless of their positions in society, in all kinds of national contexts (Ignatow et al., 2012; Johnston and Audunson, 2019; Nissen and Kann-Rasmussen, 2025; Summers and Buchanan, 2018). They also have the power to level cultural inequalities because, along with the family environment and schools, libraries provide a long-term favourable context for reading (Kraaykamp, 2003) and being taken to the library as a child is associated with upward social mobility later in life (Scherger and Savage, 2010). This multiplicity is reflected in the many definitions of the purpose of the public library, as formulated by for instance the IFLA Public Library Service Guidelines: “The primary purposes of the public library are to provide resources and services in a variety of media to meet the needs of individuals and groups for education, information and personal development including recreation and leisure. They have an important role in the development and maintenance of a democratic society by giving the individual access to a wide and varied range of knowledge, ideas and opinions” (Koontz and Gubbin, 2010, p. 2). In this sense, public libraries play an important role in the transmission and equal distribution of cultural capital (Bourdieu, 1986).

The transformation from traditional to algorithm-led collection management raises the question of how this transition will affect the capacity of public libraries to disseminate cultural capital to all sections of the public. This point of view is particularly important at a time when the use of AI in public libraries is still in its early stages and librarians are only starting to find roles and skills as AI users (Cox and Mazumdar, 2022; Cox et al., 2019). In this article, I contribute to this emerging literature through an empirical case from Finland, where some public libraries are either considering moving towards algorithm-based collection management (ACM) or have already done so. However, most libraries in Finland have not yet adopted ACM and continue to rely on other methods of collection management in which human librarians still make decisions about the acquisition, organisation and placement of collections. Based on 10 interviews with librarians at libraries either using ACM or considering moving to it, I ask the following question: How do Finnish librarians working under different collection management systems perceive the changes ACM poses for the dissemination of cultural capital? This article is structured as follows. I begin by reviewing the relevant literature on cultural capital in public libraries, the logics of algorithmic culture and recent changes in the landscape of public libraries. I then present my data and methods, followed by the findings and a concluding discussion.

ACM is built on the technology of “floating collections” in libraries – a system whereby materials are not part of any specific library's collection but rather move freely around different branches of the library network and stay where they are returned. The literature emphasises the potential of floating collections to reduce the workload of library staff and save money and resources (Bartlett, 2014). There are some, mainly descriptive, studies on the technology's practical implementations (Berg, 2022; Burman and Brage, 2016), including the case of Finland (see, e.g. Rautanen, 2016). ACM can be broadly understood as a system of floating collections that, instead of freely circulating collections whereby materials stay at the branches to which they are returned, relies on artificial intelligence to decide the library, from a wider group of branches, to which they should be sent. Finland's public libraries rely on the Intelligent Material Management Systems (IMMS), created as a collaboration between Danish company Lyngsoe Systems and Danish public libraries (City of Helsinki AI register, 2025; Liljegren, 2022). While there are already studies about how librarians tackle ACM in their work (Liljegren, 2022), there is a striking lack of research on the overarching question of whether and how the advent of AI-based floating collections affects the basic mission of libraries to provide equal access to cultural capital.

At its simplest, an algorithm can be defined as “an abstract, formalized description of a computational procedure” (Dourish, 2016, p. 3). In other words, algorithms are sequences of computational instructions or coded processes that transform data into a given predefined output. While they are “non-human technological agents” (Massanari, 2017) without ability or agency of their own, algorithms are essentially cultural; they both shape culture and are shaped by culture (Cohn, 2019). Several scholars have argued that while people know that algorithms operate behind, for instance, search engine ranking systems or different forms of digital matchmaking platforms, our everyday lives are so saturated with algorithms that they mostly go unnoticed, thus becoming invisible and intertwined with our everyday practices (Seaver, 2017). Ted Striphas (2015) coined this invisible yet revolutionary process of “sorting, classifying and hierarchizing of people, places, objects and ideas” (p. 396).

Algorithms perform a complex double role: they are mere mathematically engineered technical processes, but at the same time, their existence is a complex, socially defined and discursively negotiated and upheld process. Beer (2017) argued that the power of algorithms resides not only “in the code” but also, and essentially, in the way that they create discursive understandings of and agreements on the importance of efficiency, normality and “neutral” calculation – all desirable rationales of post-capitalist societies (see also Mager, 2012). At the same time, many scholars have noted that algorithms are not “natural, neutral, or benevolent technologies” (Cohn, 2019, p. 11); rather, they might have totally unintended effects and consequences by reinforcing and favouring the choices of the masses and excluding minorities or deviances. Scholars have shown that algorithms can contribute to the promotion of different forms of inequality, hate speech and so on by amplifying contradictory topics and voices (Massanari, 2017; Striphas, 2015).

AI is used in libraries in many ways that are not necessarily visible to its users. Cox and Mazumdar (2022) have categorized its use in libraries into five categories: (1) it is applied to routine and manual tasks, (2) it is applied directly to user services, such as via chatbots or knowledge discovery, (3) it supports larger communities of data scientists, (4) it can support information literacy and finally, (5) it can be used to analyse, predict and influence user behaviour. In other words – while libraries and librarians might find ACM helpful for tackling routine or service tasks – the library users might be rather targets of ACM's capacity for analysing and predicting user behaviour. There is very little public information available on the AI-based collection management system used in Finnish libraries (Helsingin Sanomat, 2019), but from this scant information it appears that the algorithm focuses on recognising the borrowing habits of individual libraries in a larger multi-library system by evaluating the books and other materials borrowed. As the borrowed materials are returned to libraries, the algorithmic selection process ends up re-sorting library collections based on local users' tastes, trying to make the libraries' collections respond to the kind of materials the user will be interested in borrowing. It thus seems like out of “content filtering” (meant to identify products similar in content) or “collaborative filtering” (meant to identify users with similar choices) algorithms, the ACM used in Finnish public libraries would be closer to a collaborative filtering algorithm (Cohn 2019).

The imminent challenge that ACM poses for the dissemination of cultural capital in public libraries lies in the risk that it might make collections less diverse and inclusive (Cox and Mazumdar, 2022). By “recommending” choices similar to those that others have made (whether in terms of videos, songs, clothes, media content or, in the case of public libraries, books or other materials available on the shelves), the algorithm may end up restricting, limiting or narrowing cultural repertoires and mindscapes (Beer, 2017; Gillespie, 2014; Massanari, 2017; Striphas, 2015). Thus far, empirical studies of these kinds of “filter bubbles” (Pariser, 2011) have mostly been discussed in relation to social media algorithms and their power to maintain media users in their comfort zones by boosting beliefs and political partisanship.

What is cultural capital and what can it be in the context of public libraries? Bourdieu's original framework is founded on the premise that social classes – the upper classes, the middle strata and the working or popular classes – possess unequal amounts of resources, which he conceptualised as different forms of capital: economic (e.g. money or other economical possessions), social (e.g. access to useful networks) and cultural capital (e.g. cultural knowledge or resources and skills to use for one's benefit, such as in school) (Bourdieu, 1986). According to Bourdieu's perspective, these three forms of capital are convertible. Economic capital may be transformed into cultural capital by, for instance, purchasing access to cultural events; cultural capital can be leveraged into social capital through, for example, the mastery of elite linguistic codes; and social capital can, in turn, be exchanged for economic capital by utilising one's network to secure desirable employment (cf. Reeves and de Vries, 2019).

Bourdieu further divided cultural capital into three subtypes: embodied, referring to cultural competences, such as talking or acting in the correct way; objectified, meaning the ownership of physical objects, such as books or paintings; and institutionalised, referring to officially acknowledged credentials, such as academic titles (Bourdieu, 1986, pp. 243-246). Goulding (2008, 2006\2016) famously argued that public libraries freely offer all three subtypes. Objectified cultural capital can be found in books and other materials, while embodied cultural capital can be found in the knowledge and linguistic skills these materials offer and, in turn, plays a key role in achieving institutionalised cultural capital, such as educational qualifications. It should be mentioned that Bourdieu's take on cultural capital has received ample criticism. It has been questioned whether cultural capital can ever be a fixed concept or whether its content varies according to the context (Lamont and Lareau, 1988). In recent decades, scholars have suggested that a new kind of “omnivorous” tendency to open-mindedly combine both highbrow and popular culture could replace traditional highbrow snobbery and signal high cultural capital (Peterson and Kern, 1996).

A large body of literature shows that large amounts of cultural capital are linked to high cultural participation, including active library use (Bennett et al., 2009; Heikkilä, 2024; Purhonen et al., 2019; Reeves and de Vries, 2019). Summers and Buchanan (2018) argued that low cultural capital can form many kinds of barriers to public library use that go beyond the typical (e.g. time constraints, distance and unsuitable collections). This means that people who refrain from public library use may actually feel symbolic distance towards libraries and not see themselves as potential library users or readers (Heikkilä, 2024). Even if public libraries publicly underline their role as levellers of cultural capital and are “open for all”, it is complicated to define whether “all” means that public libraries should have a neutral focus on the majority or pay special attention to underprivileged groups (Nissen and Kann-Rasmussen, 2025). This debate is reflected in what citizens think of public libraries' main role. There is a division between those who argue that public libraries should serve their real tax-paying users, particularly book readers and those who tolerantly believe that libraries should be for everyone, even when they are not used for reading (Heikkilä and Sirkka, 2023).

The advent of ACM has brought new challenges for the production, dissemination and appropriation of cultural capital. Librarians have traditionally been considered important gatekeepers of cultural capital for their local communities. They are involved in the processes of selecting and purchasing books and other materials and laying them out, thereby providing central symbolic goods and services related to cultural capital (Bourdieu, 1984, p. 359; Goulding, 2006\2016). Studies have shown that librarians are in constant negotiation between fulfilling patrons' wishes and trying to “cultivate” their tastes (Kann-Rasmussen and Balling, 2015) – in other words, keeping up with the task of disseminating cultural capital but also showing efficiency. While libraries offer many more dimensions of cultural capital than just their collections, using algorithms for these processes could introduce an entirely new logic for the traditional task of public libraries.

ACM is, of course, the result of a long continuum of different collection management styles. From traditional approaches based on professional judgment, normative visions and a certain top-down gatekeepership, libraries have gradually moved toward increasingly data-driven models that apply performance-based criteria to the selection and weeding of materials. Evans and Saponaro (2012) have outlined two main collection management trends during the last hundred years or so. First, there has been a change in the process of the collection management itself – from hands-on work with physical books, librarians now manage the physical and digital ownership of a wide range of materials and the equipment required to access them. Second, the librarian's task has evolved from gatekeeping high culture to facilitating “open access” to all kinds of materials (Evans and Saponaro, 2012, p. 34). Librarians have moved further from the material selection processes, often delegating this task to suppliers such as book wholesalers or publishers' platforms, which select the materials according to the library's specifications in the name of efficiency. ACM can be viewed as a direct continuation of these trends, prioritizing productivity and cost-effectiveness through machine learning and automatic decision-making.

Finally, it should be noted that cultural capital has been a topic of discussion in public libraries throughout their history. The so-called “fiction question” can be seen as an early manifestation of this debate. In the late 19th century, when public libraries were slowly gaining traction, fiction was considered light entertainment and a possibly unworthy genre for public library collections (Der Weduwen and Pettegree, 2021). Ellen May Schaefer, in her doctoral thesis published in 1904, summarizes the debate on whether lowbrow fiction should be available in the public libraries: “The function of the public library is educational, and it is not justified in supplying to the public flabby and worthless reading.” (Schaefer, 1904, p. 6). While currently fiction is the core of public libraries' collections, librarians still adhere strongly to the discourses of enlightment and cherishing quality (Balling et al., 2008). The task of disseminating cultural capital wisely has thus always been at the core of debates of public libraries and public librarianship – already much before Bourdieu (1984) coined this famous concept.

One of the most significant changes that public libraries are currently going through is related to the demographics of their users, which differ according to the context. For instance, in Finland, library use has recently been marked by a clear decline as well as social differentiation. Since the 1990s, the Finnish population has generally used libraries less every year, and having a high level of education and being female are significant predictors of using libraries. Library users thus match perfectly with the groups who read more books than others (Heikkilä, 2024; OKM, 2023). Similarly, research from the United Kingdom has shown that people who use libraries actively and in many different ways tend to have privileged backgrounds (Leguina et al., 2021).

On the other hand, there is evidence of libraries functioning as important “low intensive meeting places”, especially for groups with a low level of education and few resources (Aabø et al., 2010). For groups such as immigrants in particular, public libraries have shown themselves as “gracious spaces” that allow integration into society (Vårheim, 2011). This view of libraries as meeting hubs that promote social capital is heavily emphasised in legislation in Nordic countries, which now underlines the importance of libraries as promoters of democratic skills. The Finnish Library Act (2016) states the following: “The objective of public libraries and information services provided by them is to promote among citizens equal opportunities for personal development, for literary and cultural pursuits, for access to knowledge for the acquisition of personal and civic skills, for broadening their world view, and for lifelong learning.” A famous inspiration has been Jochumsen et al.’s (2012) four-space model, according to which public libraries can have overlapping inspiration, learning, meeting and performative spaces – not necessarily as physical rooms but as functions of the overall library. While it has been critically questioned how this change towards social instead of cultural functions has affected libraries' capacity to disseminate cultural capital (Heikkilä and Sirkka, 2023), it has also been noted that the mission of public libraries has not historically been born from providing cultural capital but from providing social spaces. For instance, the first proper public libraries in the United States were social and communitarian places instead of a “civilising mission” (Der Weduwen and Pettegree, 2021).

A third grand change, wholly connected to the first two, is related to the “legitimacy crisis” of libraries. Libraries not only face a decline in use and rapid digitalisation, which has forced them to rethink their activities, but they also face cuts and reductions in staff and opening hours (Guarria and Wang, 2011; Kann-Rasmussen, 2022; Robertson and McMenemy, 2018). This situation has led to a novel dilemma for libraries: in a scenario of globalisation, societal polarisation and cuts to the public sector, public libraries are being forced to compete for declining public funding by showing their importance, proving their efficiency and generally adapting to a new role as a player accepting the neoliberal rules of the game (Kann-Christensen and Andersen, 2009). In this context, library scholars have spoken about new public management (NPM) (Kann-Rasmussen, 2023). Since the 1960s, the mission of European cultural policy has largely been to level access to high culture, presenting it as a common good for everyone. However, since the democratisation of culture has proved difficult, different kinds of cultural institutions have been forced to broaden their audiences – often framing non-participation a “challenge” (Heikkilä, 2022; Stevenson et al., 2017). In this way, NPM in public libraries (and elsewhere) ends up quantifying culture instead of assessing it in terms of quality and forcing it to be instrumental (Belfiore, 2004).

These changes have led to a new focus on user orientation in public libraries. Researchers argue that a large transformation of librarians' work identities happened during the 1960s and 1970s – a period during which traditional values were largely disintegrated. This was when librarians moved from having enlightenment-minded and vocation-based identities towards emphasising patrons' experiences and their own dedication (Kann-Rasmussen and Balling, 2015). This has led to librarians being increasingly more concerned about library users' tastes and seeing themselves as forced to be very positive towards them, with librarians assumed to have a new kind of identity combining personality and professional expertise that removes the traditional gatekeeping function of librarianship and places the librarian and patron side by side as equals or collaborators (Balling et al., 2008). As librarians' work has become more user-centred, their gatekeeping of cultural capital has become more complicated. The introduction of ACM, which amplifies users' tastes over librarians' opinions, could further complicate this. How do libraries cherish cultural capital in this new scenario?

The data, which were collected for a larger research project (anonymised for review), included interviews with library managers, librarians and information specialists at two Finnish library networks as well as one week of ethnography in each of the branches. The interviews addressed the current collection management style, principles of material selection, observed changes and challenges, and the underlying ideology of libraries (see  Appendix for the entire interview questionnaire). The ethnography focused on how the library was organized as a space, how books and other materials were categorized and displayed, and what kinds of activities patrons engaged in. The library networks, anonymised here as A and B, were chosen because of their different collection management systems: A used ACM and B used floating collections, albeit with the possibility of transferring to ACM at a later stage. The libraries were located in different parts of Finland. To capture the potential inequality aspect of ACM, I first conducted interviews at the main libraries and then at individual branches from socioeconomically high and low neighbourhoods. I left this selection to the library managers, who, after the initial interviews, pointed out suitable branches and helped make contact with their personnel. The overall data are described in Table 1.

A qualitative content analysis was performed on the 10 interviews conducted in libraries A and B. I began by closely reading the interviews, looking for common topics or recurring themes. The next step was to carefully organise suitable interview excerpts under titles such as “ACM as a solution”, “resistance towards ACM” and “ownership of library work”. Finally, I coded these excerpts according to whether they expressed positive or negative stances and organised them into two larger categories: optimism and hesitation (see Chapters 4.1 and 4.2). Since only Library A used ACM, I also included a category regarding attitudes towards the imagined or expected influences of ACM (see Chapter 4.3). It should be pointed out that the units of my analysis were attitudes and discourses; this means that I did not categorise individual librarians as “optimists” or “hesitants”.

All relevant ethical principles of the Finnish National Board on Research Integrity have been followed carefully. As institutions, all libraries signed a permit to allow research on their premises and among their personnel. The research permit applications included a detailed research plan and privacy policy sheets. In addition, the interviewees were informed openly and comprehensively about the following topics: the researcher's contact information, along with detailed information about the funding of the study; the topic and objectives of the research; the means of collecting the data; the voluntary nature of participation; the provision of full confidentiality; and the fact that the data would be transcribed and anonymised for publication. To further protect the anonymity of the individual interviewees, I cite only the library to which each interviewee belongs (A1–A4 and B1–B2). It should be noted that no interviewees were available at Library B3; however, it is included in the data because other librarians at Library B discussed the differences between branches in a very lively manner.

It is only logical that when a researcher presents themself as somebody wanting to understand more profoundly how ACM has transformed the mission of public libraries to disseminate cultural capital, librarians with initially positive attitudes towards the topic will come up with their best arguments and be willing to defend themselves against the most common critiques presented against ACM. This was also true from my interviews – the majority of the interviewees with mainly favourable views had clearly thought in advance about what to say on the topic.

Perhaps their most common strategy was to appeal to the numerous points discussed in the media about how ACM actually solves many kinds of challenges present in public libraries and in workplace logistics more broadly. It was argued, for instance, that libraries suffer from a permanent situation of work overload. ACM was presented as a necessary tool to overcome it, and it was largely considered that ACM would help reduce the manual and repetitive parts of work, presented as dull or uninteresting.

While logistics certainly form a core part of the pro-ACM arguments, another central part is tech-positive ideas about AI in general being a progressive, benevolent tool that essentially does something much better than humans. Often, this was expressed through the ideal of serving patrons. In the following discourse, it is genuinely believed that ACM is more successful in knowing what patrons want:

This belief typically receives support from the idea of ACM being more democratic and neutral than humans in understanding what patrons want – essentially because “the numbers know”. The librarians often argued that ACM was more democratic than traditional collection management because it took power from individual librarians and scattered it to larger teams who consulted the “wise” flow of numbers.

Another reasoning directly related to patrons is the often-mentioned fact that ACM is linked to the greater circulation and turnover of materials. It was often mentioned that this is something patrons had been happy with – ACM had made their branch libraries' collections look different every time they visited them and had brought novelties more quickly to their reach. Many interviewees defended ACM as a way of sensing trends and, again, bringing to patrons something that perhaps would be beyond the librarians' tastes. This was well formulated by an interviewee who speculated that librarians, with their canonical and old-fashioned literature tastes, might not even be able to recognise the popular literature that patrons actually desired:

The tendency of ACM as a technology to know what patrons want was praised as a way of making the collection more efficient. This made it easier to decide to acquire less “marginal stuff” – in other words, seldomly loaned or more niche materials – as the following interviewee put it:

Some of the librarians recognised that with the use of ACM, the traditional image of a public library changes. In the interviews, it was described how libraries undergo an inevitable transformation from an encyclopaedic, collection-orientated milieu into a highly functional minimalistic space, based on what the patrons want.

Finally, it should be emphasised that while many of the librarians felt they did not know ACM well enough to feel a certain ownership towards it, as we shall see in the next section, those positive towards ACM underlined that it is a technology that should always be used together with people – in their case, with professionals – not as a faceless machine that is allowed to roam on its own.

As we have seen in the optimism discourse, there was a strong belief that ACM is, in general, a helpful and positive transformation in libraries. While there were no outspoken concerns about cultural capital, those in the optimist discourse seemed to believe that cultural capital is connected to how well ACM recognises what patrons want. Some librarians saw that ACM was able to discern the tastes of patrons better than librarians and thus serve them efficiently. Overall, the optimism discourse is perhaps best understood in the context of the constant pressures of the public sector to justify the funding they receive (Belfiore and Bennett, 2010; Dubois, 2015). Libraries have to show that they are efficient, and the transfer to ACM as a vehicle for better curating patrons' tastes and taking the best out of the collection serves this purpose perfectly. In the end, it could be argued that ACM maximises the potential of quantified and instrumental public library use, particularly when it comes to book circulation. The argument seems to be that high lending numbers are cultural capital – they are portrayed as a success of ACM. We next see a very different discourse – that of hesitation.

Some interviewees spoke about cultural capital with an opposite twist, believing that ACM has largely undermined the possibilities of disseminating cultural capital in public libraries. First, it was seen to have weakened the ownership of what used to be the librarians' “own” libraries and collections, making book selection more anonymous and scaling acquisition decisions to higher levels and to “numbers”. Some of the librarians wistfully reminisced about the times when they felt they still controlled their collections themselves.

For many librarians, ACM essentially meant a strong resistance to change. At this point, some of them argued that it had taken from them the parts of their work that they liked the most and the best of their professional skills, particularly the selection of books. As one librarian put it,

For some of the interviewees critical of ACM, the criticism was directed towards a feeling of not properly understanding what this new technology actually does. On this point, even the interviewees fairly favourable towards ACM agreed that the processes of “analysing, reporting, compiling statistics, training the employees is still a bit underway”, as one of the highly tech-positive librarians at A1 stated. The more critical interviewees tended to phrase this as a way of, again, taking ownership of work away from the librarians themselves:

Thus far, the criticism in the hesitation category has mostly revolved around logistics – the fact that ACM had to a point swallowed tasks that were considered professionally interesting for the librarians and that, at the same time, its operating logic was not clear to them. Another large stream of criticism and resistance lies in a more substantial argument about cultural capital: as ACM mainly focuses on how books and other materials are lent out to patrons, it quickly detects which materials are not circulating as efficiently as others and rapidly suggests weeding them. In this way, ACM was often presented as a destructive counterforce to cultural capital.

Apart from the weeding, the librarians were worried that ACM would lead to libraries providing patrons only with materials close to their existing tastes and not nudging them to, for instance, take detours and expose them to new kinds of materials they were not looking for. This, according to many librarians who were critical of ACM, was linked to the weakening of the core “civilising mission” of libraries.

We have seen that in the hesitant discourse, there was criticism of ACM from two main points of view: ACM was considered (1) to weaken librarians' ownership of meaningful tasks and, eventually, the collections and (2) to make the idea of encyclopaedic public libraries dangerously redundant. The hesitants often encapsulated the idea of algorithms being “black boxes” (Gillespie, 2014; Pasquale, 2015); some of the librarians mentioned that they did not really know how ACM works, even if it is responsible for how books and other materials travel around the different branches. They were also worried that ACM would eventually limit and narrow not only patrons' cultural repertoires but also entire collections, which has been a central argument in the literature on algorithms (Gillespie, 2014; Massanari, 2017; Striphas, 2015). In this sense, the hesitation discourse was close to the Bourdieusian idea of the existence of objectified cultural capital residing in the acquisition of high-quality cultural materials (Bourdieu, 1984). Central to the hesitation discourse was also the belief that public libraries are capable of levelling cultural capital between citizens (Goulding, 2008; Summers and Buchanan, 2018). Next, we see which kind of discourse was closest to the librarians whose libraries did not yet have ACM.

Only Library A actually used ACM, while Library B had floating collections – a prerequisite for ACM – and had considered transferring to a new, algorithm-driven collection management system. Perhaps surprisingly, the attitudes of the librarians in Library B were closely related to what we had already seen in the optimistic and hesitant comments, and they, too, revolved around the topics of work logistics, on the one hand, and collections and patrons, on the other hand.

The first topic – logistics – was discussed mainly through excitement and envy. From the point of view of a library system with floating collections managed manually, ACM looked like something that could ease tedious everyday work tasks. The librarians at Library B spoke abundantly about how hard it was to control the flow of books typically lent out in one branch and returned to a different one. In fact, most of the interviewees at library B mentioned branch B2, located in the highest socioeconomic area of the city, which had the constant problem of accumulating not only the most prestigious fiction but also specific categories of non-fiction close to the tastes of the area's wealthy residents – for instance, finance and economics. This made the books radically different from those accumulating at B3. In the following quote, ACM is portrayed as a possible solution:

Another logistical improvement that the librarians who did not use ACM discussed in a highly positive way was obtaining real-time information about the circulation of the collection. Some librarians felt that it would be important to gather finer-grained data about how the materials moved around, not only to ease logistics for themselves but also to make the materials better suit the patrons.

From this comment, it becomes clear that the librarian is interested in ACM not only to organise the workflow more efficiently but also to serve patrons better. While the positive ideas about ACM usually revolved around technical solutions that would eliminate work tasks considered boring or repetitive, the negative imaginary concentrated entirely on possible threats that ACM could introduce in terms of limiting or narrowing the selection of materials available for patrons, thereby narrowing their entire worldviews. In the following, the interviewee makes references to the diverging cultural and social resources that patrons have, assuming that not all of them would benefit from ACM in the same way:

Another highly recurrent theme is the possibility that ACM would make libraries' collections more homogeneous and therefore be less likely to take patrons out of their comfort zone – something that was very highly valued by all the librarians, including most of those already working with ACM. A librarian in charge of acquisitions at Library B mentioned Enni Mustonen, one of Finland's best-selling authors, with more than 100 works of popular historical fiction, and considered it a risk that ACM would push libraries into emphasising popular, already widely circulating materials, such as Mustonen's bestsellers:

Finally, studying the expected influences of ACM showed that librarians who worked under different kinds of collection management had many opinions and ideas about it. Interestingly, this discourse seemed to combine elements from both the optimism and hesitation discourses. While it was seen as positive for library professionals that ACM would simplify logistics and remove redundant work tasks, it was considered a risk that libraries' collections would become more limited and unsurprising, thus ceasing to surprise or educate patrons. The concluding discussion pays more attention to this tension.

In this article, I set out to ask how Finnish librarians working under different collection management systems perceived the changes posed by ACM for the dissemination of cultural capital. We saw that the librarians had highly diverging views, which I categorised as optimism and hesitation. In the optimism discourse, it was understood that ACM was an improvement for work logistics and as a neutral, benevolent gatekeeper able to surpass the often fuzzy and biased logics of humans. Meanwhile, the hesitation discourse identified important challenges regarding the diversity of collections and lamented that ACM had deprived librarians of professional ownership of their collections. These diverging discourses had highly differing interpretations of how ACM affects cultural capital in public libraries. The optimism discourse tied it to ideals of efficiency and pleasing patrons, whereas the hesitation discourse regarded cultural capital as something that grows from exposure to rich, diverse materials in public libraries (see Table 2 for a summary). While the discourses were highly different, what they had in common was that both groups of librarians had an enormous will to market their “product” – that is, the books and other materials available in their libraries (Kann-Christensen and Andersen, 2009). Both discourses are presented and justified as a means to encourage patrons to use libraries and to read more.

The two diverging viewpoints can be better understood by contrasting them with the much-discussed public administration reforms – NPM and the more recent New Public Governance (NPG) – both of which have been extremely influential in the public sector, including libraries (Kann-Rasmussen, 2023). While NPM, introduced in the 1970s and 1980s, focused on incorporating business administration principles, such as measurement and assessment, into the public sector, NPG, emerging a decade or two later, began to emphasise ideals of participation, values and multiple stakeholders, casting public administration workers as active, self-regulating and collaboration-orientated players in the field. Kann-Rasmussen (2023) argued that despite its emphasis on economic efficiency, NPM still allows librarians to retain the core of their professional identity – that is, direct contact with patrons, expertise regarding materials and so on. In light of this, my findings on optimism versus hesitation can be understood as a negotiation between the ideals of NPG and NPM. The hesitation discourse tolerates the efficiency principle of NPM but has trouble accepting the multiple stakeholder ideals of NPG. Meanwhile, the optimism discourse has moved towards believing in NPG's goals and has embraced the ideals of innovation potential and trusting that cultural actors can promote positive change in the context of public institutions, such as libraries (Kann-Rasmussen, 2023).

How can my findings be interpreted in the light of the debates on cultural capital? An evident limitation of this study is that I only examine materials available in the libraries as a proxy for cultural capital – although, of course, it is also fostered in many other contexts. Bourdieu originally introduced the concept of cultural capital to explain why children from educated families performed better at school than others. According to him, formal school curricula included items – such as highbrow arts or books – that were already familiar to children from educated families, who then were recognized and rewarded for something socially inherited rather than learned at school (Bourdieu and Passeron, 1979, Lareau and Weininger, 2003). As Goulding (2008) and others have argued, public libraries have the capacity to break this vicious circle by levelling the threshold for acquiring cultural capital. In this sense, ACM poses a genuine challenge to the public library's task of disseminating cultural capital by openly questioning the relevance of gatekeeper-librarians and the need for large collections, including also slowly circulating materials such as highbrow and niche literature. In other words, the traditional understanding of disseminating and levelling cultural capital might be at risk if ACM increasingly takes over the selection, curation and weeding of materials. Nevertheless, scholars agree that book collections are not the only means through which libraries offer cultural capital. This is also achieved, for instance, through addressing social exclusion and fostering communities (Goulding, 2008), by creating new kinds of more personal recommendation models to spark interest in books (Loh et al., 2022), and by tackling digital inequalities more efficiently (Leguina et al., 2021).

To put my findings into a broader context, the use of ACM is just one drop in the ocean of cultural transformations – in particular, the decline of the status of highbrow culture and the gradual legitimisation of popular culture (DiMaggio and Mukhtar, 2004; Purhonen et al., 2019), the financial pressures of the public sector (Johnson, 2018; Robertson and McMenemy, 2018) that have pushed libraries and other public institutions towards ideals of efficiency (Belfiore and Bennett, 2010; Day, 2002; Kann-Rasmussen, 2022) and the ubiquity of algorithms in libraries (Cox and Mazumdar, 2022; Cox et al., 2019) and society at large (Striphas, 2015). While ACM's logistical advantages are clear, it is possible that its combination with ongoing financial austerity measures and persistent pressure on public libraries to demonstrate their usefulness and efficiency could lead to easy justifications for reducing the breadth of public library collections or making further cuts in personnel or library spaces. In a scenario of declining and stratified library use, it is important that library professionals, policymakers and library scholars are aware of how technological changes such as ACM can amend one of the core missions of public libraries – which is to create, disseminate and level citizens' cultural capital.

I wish to thank Koneen Säätiö for their financial support of the research project, my informants for their insight and time, and the reviewers of this article for excellent and helpful comments.

1. When did you start with the current collection management style?

2. How was the information on the current collection management style received within the community of librarians: what kinds of actors were involved, how was the process, what costs did it include for the library?

3. Was this library among the early or late adopters of this collection management style?

4. Can you tell me something about the implementation and schedule of the current collection management style?

5. How does it work, technically: is there information available about the development and operation of the collection management style? Who is in charge of monitoring it?

6. Could you tell me about the advantages and disadvantages of the current collection management style?

7. Are there concrete consequences of the current collection management style: have any changes been made based on them?

8. In what ways is the current collection management style visible at this library?

1. On what basis are materials acquired for the library?

2. On what basis is the collection curated (acquisitions, maintenance, weeding)?

3. On what basis does the collection combine so-called “high-quality”, “artistic”, “popular”, “commercial”, “entertaining”, etc. materials?

4. On what basis are books displayed and shelved in this library?

1. What kinds of everyday challenges are there at this library?

2. What kinds of things beyond reading are available at this library?

3. There has been a lot of talk in research about libraries becoming concentrated on experiences: how do you see this?

4. Research has shown that reading is declining and library use is decreasing: what is the reason, from the perspective of your own expertise?

5. What dangers or threats do you see in the crisis of reading or the use of libraries?

1. From the perspective of your own expertise: for whom are libraries made?

2. What is the role of a public library as a maintainer or developer of cultural taste, even cultural capital?