Introduction
Individuals often perpetuate inequalities as a result of their attitudes about what is valuable or normal within their communities. Philosophers are no exception: whether implicitly or explicitly, many of them have taken certain aspects of the discipline as normal or as more or less valuable within it. For instance, some topics have been treated as core and others as fringe, some traditions and methodologies have been treated as less central or not philosophical, and even some individuals have been treated as more or less worthy of the title 'philosopher.' But it stands to be said that where these attitudes are present, there are often good grounds for worrying that prejudicial, paternalistic, or intellectually careless or even vicious attitudes underlie them and that highly problematic consequences will likely follow from them, including unwarranted silencing, marginalization or exclusion, tokenism, paternalism, objectification, systematic misrepresentation, and even exotification. In our understanding of these attitudes and dispositions and their consequences, we are borrowing from the various literatures on epistemic violence, oppression, injustice, and viciousness as discussed by, among others, Collins (Reference Collins1990); Mignolo (Reference Mignolo2000); Fricker (Reference Fricker2007); Dotson (Reference Dotson2011, Reference Dotson2012, Reference Dotson2014); Cassam (Reference Cassam2019); and Kidd (Reference Kidd, Kidd, Battaly and Cassam2021).
In ‘Categorical Imperfections: Marginalisation and Scholarship Indexing Systems’, Simon Fokt (Reference Fokt2020) highlights that subject indexing is one area where such attitudes and dispositions seem to arise and result in the sort of problematic consequences just mentioned. Using the example of PhilPapers, the most commonly used Anglophone indexing system in philosophy, Fokt shows that ‘subject indexing is a “communicative practice” . . . which uses discoverability and subject representation . . . to carry implicit judgements about the relative importance, generality, and complexity of the subjects it includes’ (Fokt Reference Fokt2020: 220). In this context, it is important to recognize and revise ways in which content falling outside what has traditionally been perceived as the core of Anglophone analytic philosophy has been represented in such a way as to suggest that it is less valuable, less complex, and even less philosophical, when arguably it is not.
For instance, PhilPapers’ ‘Browse by Category’ groups all world philosophical traditions—except the Western analytic tradition—in the crypto-miscellaneous category Philosophical Traditions, which appears toward the end of the nonalphabetical list of top-level categories (PhilPapers n.d.). It is not divided in a careful or meaningful way, and its specific topics are found lower down the category tree and are given tradition-specific qualifiers that distinguish them from unqualified (and thus presumably universal) categories where Western analytic scholarship is found. Similarly, feminist perspectives tend to appear in miscellaneous categories, with qualifiers distinguishing them from unqualified scholarship (for example, Feminist Philosophy of Science as a subcategory of Philosophy of Science, Misc.).
Moreover, when it comes to topics, the ordering of top-level categories on PhilPapers strongly suggests a gradation of importance of philosophical topics, with those more abstract and, as it happens, more dominated by white male authors appearing closer to the top. Categories devoted to times, figures, and topics more relevant to the Western analytic tradition tend to appear higher up the tree, have more subcategories, and lack tradition-specific qualifiers. Scholarship discussing other times, figures, and topics is listed under (crypto-)miscellaneous or tradition-specific categories and is often not found in the general, unqualified categories.
Overall, the popularity of PhilPapers, the fact that it is co-created and maintained by members of the philosophical community, and the fact that there is little critical discussion of these facets of the database all indicate that the way it represents the discipline has simply been accepted as normal. The result is a transparency effect—the way PhilPapers (and other resources) represents the discipline in their categorizing of it will lead philosophers to interpret this representation as the way the discipline actually is, with some parts being more or less important or complex. Following Sally Haslanger (Reference Haslanger2012), perhaps, Western analytic philosophers’ ideologies concerning how they practice philosophy has become so hegemonic that they ultimately cannot disconnect how they practice philosophy from philosophy itself without an extreme stretch of their imaginations.
Naturally, accurately representing the various areas within a discipline as more or less complex, central, or general is a desirable feature for any indexing system. But problems arise when the representations embed illegitimate attitudes and dispositions. For example, if the work of Confucius is in fact roughly as complex and central as the work of Plato, then a system in which the category Confucius is represented as less important, complex, central, and so on, than the category Plato is problematic—especially, in the Anglophone context where Chinese philosophy is often marginalized. Such a system marginalizes the category Confucius and consequently marginalizes and is unfair to the work of Confucius and the many scholars associated with his work. Equally, it misrepresents the discipline.
Further, authors from marginalized groups are more likely to research topics and traditions that are marginalized by many indexing systems, including PhilPapers (Botts et al. Reference Botts, Bright, Cherry, Mallarangeng and Spencer2014). Hidden in categories that are presented as less important, less central, less valuable, their work might be less often read, less often discussed, less often remembered. And listed in the mixed bag of miscellaneous topics, it might also imply that the authors lack intellectual coherence, are unimportant, random, or unfocused when they are not. Accordingly, these authors face the very high likelihood of being subject to the problematic consequences mentioned above.
We take such consequences to be a problem for any resource that uses an indexing system—especially when it claims or implies to be relevant to all of philosophy. Naturally, a resource focused on Anglophone analytic philosophy, or on metaphysics and epistemology, might be right to make its stated core topics more central and, accordingly, represent them as more important. For instance, a resource named AngloAnalyticPhilPapers would clearly indicate such a focus (Botts et al. Reference Botts, Bright, Cherry, Mallarangeng and Spencer2014: 231). That being said, resources might be right in renaming themselves and being more restricted in content, but contrary to our purposes, this strategy might result in unwarranted epistemic bubbles (epistemic structures that simply exclude by omission) or echo chambers (epistemic structures that cast those outside of the structure as epistemically untrustworthy or misguided). Both sorts of structures could undermine the exchange of valuable epistemic resources between those within the discipline. But if vicious enough, echo chambers could also result in contempt, hostility, or perpetuate ignorance (Nguyen Reference Nguyen2018: 141–49).
And resources that retain this anglo-analytic focus but claim coverage of all of philosophy would rightly be perceived as illegitimately exclusionary because they wrongfully deny equal epistemic status and respect to those who deserve it and, consequently, limit discussion and available epistemic resources. Equally, in willfully denying any need for change and only allowing the use of structurally prejudiced or exclusionary resources, it can also be argued that a form of contributory injustice, as set out by Kristie Dotson (Reference Dotson2012), will be perpetuated, and, in effect, various philosophers will be ignored or compromised, given their use of alternative resources, and as a result will be thwarted in their ability to contribute to the philosophical community (Dotson Reference Dotson2012: 32).
Fortunately, in recent years, such issues have been coming to light, and thus exposed, they have been waning as many (Anglophone analytic) philosophers have begun to recognize the unfair treatment of previously marginalized people, traditions, perspectives, and topics. Accordingly, we can also see an expanding number of voices and initiatives calling for and promoting a fairer and more inclusive approach to the discipline. We believe that, as the views commonly held within the community are changing, it is also high time that the indexing systems commonly used for classifying our discipline's work started to reflect that change. Indeed, Following Mignolo (Reference Mignolo2000) and Quijano (Reference Quijano2007), among many others, our project can be viewed as a part of the overarching project to decolonize or unlink philosophy from its past and present reality as a discipline primarily practiced by affluent white males from the English-speaking Global North.
No doubt, the primary obstacle to change is inertia. Change is difficult. PhilPapers alone has more than 2.5 million entries, and reclassifying all of them would be a titanic task. So, although those who manage PhilPapers and other databases might want to revise their indexing systems, it is hardly surprising that they might be reluctant to implement any substantive changes until a working model is established. And so, to accommodate this need, we offer a model indexing system for the Diversity Reading List in Philosophy that will avoid various sorts of marginalization and exotification and thus represent the discipline's work as fairly and inclusively as possible while at the same time retaining or even improving search and browsing usability.
The general mission of the Diversity Reading List in Philosophy is to help people make a positive change by reducing the amount of work and time required to achieve it. We see our proposal as another step on that mission and hope that our model will become the start of a discussion that will facilitate practical changes in the indexing systems of various philosophical databases in the near future.
1. The Diversity Reading List in Philosophy: Its Mission and Current Structure
There is a great deal of evidence of gender and racial inequality within academic Anglophone philosophy, and numerous concerns have been raised about where the discipline is centered.
For example, within academic Anglophone philosophy, gender disparity remains a significant problem. In the United States, women constitute only 25 percent of US Philosophical Gourmet Report–ranked faculty, 29 percent of recently placed philosophy PhDs and of recently graduated philosophy PhDs, and 24 percent of APA members who reported their gender (Schwitzgebel and Jennings Reference Schwitzgebel and Jennings2017). Things are only slightly better in the United Kingdom, where in 2021 women constituted 30 percent of permanent staff (up from 24 percent in 2011), 33 percent of recently placed philosophy PhDs (up from 31 percent in 2011), and 32 percent of recently graduated PhDs; still, the drop-off from undergraduate to Master's and PhD-levels has remained the same at -15 percent (Beebe and Saul Reference Beebe and Saul2021: 9). Meanwhile, only 13 percent of authors published in top American and British philosophical journals are women (Schwitzgebel Reference Schwitzgebel2015a, Reference Schwitzgebel2015b). Racial and ethnic disparities are even more pronounced. None of the PhDs graduated in the United States in 2017–2019 were Indigenous Americans or Alaskan Natives, 3 percent were Black or African American, 4 percent were Asian, 6 percent were Hispanic, and 3 percent were either more than one race or an unrecognized race (Schwitzgebel et al. Reference Schwitzgebel, Bright, Jennings, Thompson and Winsberg2021). Further, only 0.32 percent of US authors of research publications in top philosophy journals are black (Bright Reference Bright2016), and only 1.32 percent of philosophy faculty are black (Botts et al., Reference Botts, Bright, Cherry, Mallarangeng and Spencer2014). No systematic data is currently available for the United Kingdom. But what is obvious is that there are less than 1 percent permanent Black philosophy faculty members in the United Kingdom (Coleman Reference Coleman2014) and, more specifically, there appear to be only six (Tremain Reference Tremain2019). Academic Anglophone philosophy in its conceptions of philosophy has also been accused of being too Anglophone-centric, Global North-centric, ableist, class-biased, and more; see, among others, Tremain (Reference Tremain2014); Dussel (Reference Dussel and Demenchonok2015); Chiesa and Galeotti (Reference Chiesa and Galeotti2018); De Cruz (Reference De Cruz2018).
To help mitigate these inequalities and worries, the Diversity Reading List in Philosophy provides more visibility and access to works by authors from under-represented groups and makes it easier for educators and researchers alike to include those works in their respective practices.
The Diversity Reading List in Philosophy acknowledges that diversifying a curriculum (as well as one's own research) is time-consuming, since the relevant works are less likely to be popular or available, and finding them and assessing their relevance and usefulness often involves considerable effort, adding to the busy schedules of educators and researchers. To reduce those barriers, the Diversity Reading List in Philosophy (1) preselects relevant work: its database includes works that have been recommended by philosophers, assessed by a team of specialists in the relevant areas, and included based on their academic quality and relevance to current teaching; and (2) aids the final selection: texts are classified based on how specialized and readable they are, and short comments help educators and researchers quickly sieve through the content to find what they need. As a result, those who use the Diversity Reading List in Philosophy can diversify their teaching and research considerably more quickly and easily. Further, as the number of users grows, so will the number of more diverse course materials and courses provided to students who, in turn, we hope, will come to build a fairer and more inclusive discipline. Although the Diversity Reading List in Philosophy's mission has not been formulated in terms of a particular, preexisting pedagogical approach, we are certainly sympathetic to a number of them, including critical approaches (Freire Reference Freire1996; hooks Reference hooks1994), and multicultural approaches (Banks Reference Banks2016).
As a volunteer-run project, the Diversity Reading List in Philosophy did not have the resources to develop a brand-new indexing system for its entries. Instead, it adapted the existing PhilPapers category structure with some minor modifications. PhilPapers has maintained seven main topics or categories presented in the following order (numbers of entries in parentheses):
1. Metaphysics and Epistemology (405,913)
2. Value Theory (626,956)
3. Science, Logic, and Mathematics (485,348)
4. History of Western Philosophy (364,660)
5. Philosophical Traditions (278,056)
6. Philosophy, Misc. (7,997)
7. Other Academic Areas (101,862)
At the time of writing, the Diversity Reading List in Philosophy had six main topics or categories:
1. Value Theory (739)
2. Metaphysics and Epistemology (514)
3. Science, Logic, and Mathematics (263)
4. History of Western Philosophy (127)
5. Philosophical Traditions (124)
6. Philosophical Education (26)
PhilPapers currently has a total of 5,827 further topics or categories.
The Diversity Reading List in Philosophy's main categories are ordered by the number of entries within them, going from greatest to least, while the PhilPapers ordering seems to follow traditional views on the relative importance of different areas. The Diversity Reading List in Philosophy further omits the miscellaneous categories, retaining only Philosophical Education, which on PhilPapers is found within Philosophy, Misc. In both resources, main categories include subcategories (such as topics, subtopics, geographical traditions, individuals, and so on), which are ordered alphabetically or chronologically. Figure 1 displays both a topical subdivision from the Diversity Reading List in Philosophy's main category Metaphysics and Epistemology, which is ordered alphabetically; and a subdivision from its main category History of Western Philosophy, which is ordered chronologically.
Indexing subdivisions from the Diversity Reading List in Philosophy
An alternative way to browse the list is by using a word cloud of philosophical keywords (such as aesthetics and autonomy; figure 2). Users can also use a general search function, and filter search results by category, medium, recommended use, and difficulty, thus narrowing the search to, for example, easy to read, introductory book chapters and journal articles about time travel within the category Metaphysics and Epistemology (figure 3).
Word cloud of keywords, Diversity Reading List in Philosophy
Search function with filters, Diversity Reading List in Philosophy
2. The Big Picture
The primary point of an index for a database is not always to provide users with a straightforward direction to the specific resource for which they are looking. A person who knows exactly what they need will be much better served by simply using a search function. An index, on the other hand, will primarily serve the needs of those who know only roughly what they are looking for, who want to find resources related to ones they already know, or who want to get an overview of a topic. It is in such exploratory research that a category-tree structure is most useful. This puts more stress on editors to ensure that their category structures do not make it harder to find traditionally marginalized content.
We recognize that it is not practically possible to develop an ideal system that would represent the discipline without any distortions. Firstly, there are limitations inherent in any indexing system with a category-tree structure. Category-tree structures with higher-levels and lower-levels, for example, cannot easily represent cases where a single lower-level category is related to multiple higher-level categories or, similarly, to lower-level categories of other higher-level categories. For example, the more specific topic Aesthetic Experience might be best categorized in both the higher-level category of Aesthetics and that of Philosophy of Mind.
Secondly, an indexing system must marry fairness and inclusivity with functionality. Making a system perfectly inclusive would likely make it extremely complex and alienate users. For example, it might be possible to circumvent the above-mentioned limitations of a top- and sub-level category structure by adopting an interactive 3D bubble graph of interconnected topics. But while such a graph might be interesting to look at and informative in mind-mapping the discipline, it would likely not be very useful to people who want to browse a database with ease.
Thirdly, fairness and inclusivity must also be married with ease of editing. More detailed searching, browsing, and filtering might be welcome by users, but, practically speaking, editors must be able to implement and maintain their databases without overly burdensome workloads.
And lastly, for many indexing systems, language is going to be a major limitation on inclusivity. For both the Diversity Reading List in Philosophy and PhilPapers, for example, English is the predominant language both in the display of categories and entries and in how they are contributed (although PhilPapers does have a growing number of non-English entries, which is a step in a the right direction). Ideally, there would be a fully inclusive database for all philosophical work in any language that could accommodate any potential user and represent any potential author. However, editing such a database would be immensely time-consuming and would require a large number of editors who would also need to have both the requisite language abilities and philosophical understanding. And so, although we enjoy the idea of such a database, it is unfortunately not one that can, practically speaking, be implemented as of yet.
But even if it is not practically possible to create a perfectly fair and inclusive indexing system that will represent the discipline without distortion, we can and should try to develop a system that is as fair and inclusive as possible, given the concerns that we have sketched in our introduction. Below we present several general requirements that we believe any indexing system aiming to represent philosophy fairly and inclusively should meet, based on the lists of problems and proposed solutions offered by Fokt (Reference Fokt2020). We also describe more specifically features that we hope will be implemented by the Diversity Reading List in Philosophy's indexing system in particular, and databases in Philosophy more generally.
2.1 Valueless Ordering of Categories
Orderings can be used to indicate the relative importance of categories. For example, categories appearing first can be taken as more important or more general. In principle, a good category structure should represent categories as more important or general only if they are, in fact, so. All other cases call for valueless ordering—such as alphabetical, chronological, or some other order that does not imply differences in importance or generality.
Within the Diversity Reading List in Philosophy, the top-level categories are ordered from greatest to least in terms of the number of works included within them. While this was envisaged as an improvement on the value-laden PhilPapers ordering, we advise against it because when the popularity of a topic or the prominence of a geographical region determines what is prioritized, the result is often marginalization of thinkers based on what topics or regions they consider (Fokt Reference Fokt2020: 234). Instead, alphabetical ordering should be preferred for both main and subcategories, as it is value-neutral and most intuitive for users.
2.2 Equal Depth for Equal Importance
The level at which a category is placed can imply its generality and relative importance, with top-level categories perceived as being most general and likely most important, and each consecutive subcategory level being less so. A good category structure should ensure that categories of similar generality are listed at a similar level. In particular, it should ensure that categories where marginalized content is likely to appear are not listed deeper down the tree, thus implying that they are less general and important.
For instance, History of Western Philosophy is a main category in PhilPapers and, derivatively, the Diversity Reading List in Philosophy. But the historical categories from or about other philosophical traditions, they are all sub-sub-categories within the main category of Philosophical Traditions. This result strikes us as unacceptable. It implies that the history of Western philosophy is more important and has more depth than other historical traditions—which is not the case. Further, with the right sort of presentation, we might also avoid cases where categories such as Arabic and Islamic Philosophy prioritize an Arabic ethnic identity or the Arabic language over, say, Persian ethnic identity or the Persian language. Similarly, we can also avoid categories like Asian Philosophy, which are simply far too coarse-grained, containing a largely unstructured collection of thinkers from very different times and places, lumped together into a single, unstructured category.
2.3 Equal Division for Equal Complexity
Relatedly, categories with more subcategories will likely appear richer, more important, and more complex. So, to avoid misrepresenting some topics as more or less rich, important, or complex than they really are, a good category structure should divide categories into a number of subcategories that is roughly proportional to the richness, importance, and complexity of their content. In particular, it should ensure that categories where marginalized content is more likely to appear do not feature proportionally fewer fine-grained subcategories than those covering content traditionally considered to be 'core.' For example, Philosophy of Mind might subdivide into ten lower-level categories—but since Aesthetics can also do so, it should also subdivide into ten lower-level categories, not merely three or four.
2.4 Decentralized Naming
Including a contextualizing qualifier for a category can suggest that it is more specific, while a lack of such qualifiers can suggest centrality, generality, and universality. A main category named Ethics will seem general, with users expecting to find content that is central to the topic. Meanwhile, categories such as Chinese Ethics, Feminism: Ethics, or Ancient Ethics will seem more specific, fringe, and less central to ethics in general—especially, if they are displayed as sublevel categories. An index should qualify categories only where it is in fact useful and meaningful and, equally, should avoid placing marginalized content in qualified categories, where it will likely continue to be perceived as less central, general, and universal, and remain less likely to be found by people browsing the topic. If anything, content classified under qualified categories should also be found in a respective unqualified category.
2.5 Careful or No Use of Miscellaneous Categories
Content placed in categories labeled Miscellaneous or Other or categories not explicitly labeled as such but used in a similar way in practice, can seem less important or central to the issues covered within top-level categories. A good category structure should ensure that only content that is, in fact, fringe or difficult to classify otherwise, is found in such categories. In particular, it should ensure that marginalized content that is not difficult to classify otherwise, is not relegated to such categories.
For the Diversity Reading List in Philosophy, we think that Miscellaneous categories can be done without, seeing as most, if not all, philosophical works are to some extent miscellaneous in content, thus making a principled demarcation difficult to justify in such a way as to avoid the risk of potentially marginalizing some philosophers’ work. On the practical front, we believe that removing them would improve an index's usability, seeing as many users will likely find it easier to identify and search through a predictably related category than it is to search through an entire Miscellaneous category. Granted, this decision might seem as if we are boxing philosophical works into categories. But, we do not see a substantive issue here, considering all works will be allowed to be multiply categorized.
2.6 Summation
Our overall aim is to design an indexing system that can inclusively and fairly represent as much of philosophy as is practically possible. In PhilPapers and, derivatively, the Diversity Reading List in Philosophy's current systems, topics traditionally considered 'core' have the appearance of being more important, central, and complex than others, and Anglophone analytic philosophy has the appearance of being philosophy, while other traditions are somewhat exotified and marginalized. The abovementioned aims may not be sufficient for alleviating all the worries that might be posed about various indexing systems, but we think that they will help to steer us in the right direction. In combination with the modifications that we will now turn to proposing, we submit that a better balance can ultimately be achieved in terms of fairness, inclusivity, and user efficiency for both the Diversity Reading List in Philosophy and various other databases of philosophical works.
3. A New Structure
Although no one indexing system is going to be perfect, some of them can be better or worse when it comes to user efficiency and representing the discipline fairly and inclusively.
Broadly, our suggestion is to retain an expansive category-tree structure covering different alphabetically ordered topics that, unlike the current PhilPapers and Diversity Reading List in Philosophy category structures, would not include categories covering different figures, times or traditions. Rather, we suggest that selecting a topic should display all content on that topic, and time, traditions, and figures should be filters that will allow users to narrow the results.
There are two steps to our proposal. The first is a general structure for the index—the division between topics, time, traditions, and figures. The second is the specific content—the specific categories, traditions, and criteria for figures that we suggest. At this stage, we primarily defend the former, although we include suggestions for the latter in order to encourage further discussion.
3.1 Topics (What?)
Although existing category-tree structures can currently serve a number of purposes, most of them still require substantive re-prioritizing. For instance, historically, some topics have been deprioritized (or excluded) and others prioritized as a result of what Western analytic philosophers have found to be central, engaging, or important. For example, in the PhilPapers index, Metaphysics is featured prominently as a first-level subcategory of a main category Metaphysics and Epistemology, which is displayed first (out of alphabetical order) on the list, whereas Feminist Metaphysics is displayed as a second level subcategory, under Philosophy of Gender, Race, and Sexuality—a catch-all crypto-miscellaneous category where it does not really fit, and which is displayed under Value Theory, where it also does not necessarily fit.
The general structure of the index that we propose involves (1) including a broader range of topics as top-level categories; (2) limiting this range to a practically usable number; (3) removing miscellaneous and figure-, tradition-, or time-based categories; and (4) arranging all categories at all levels alphabetically.
The specific structure of the index, such as which specific categories appear at the top level, should be relative to the needs of the particular resource database. Some databases might be dedicated to a specific area of philosophy that editors may or may not want to feature more prominently; or they might be aimed at less experienced users who need a narrower range of options to grasp. Further, there will likely be reasonable disagreement within the discipline as to how topics should be presented. Below, we outline a structure we believe adequate for the Diversity Reading List in Philosophy, which might serve as a starting pointing for future discussion.
Firstly, we see the top-level categories and first sublevel categories used by PhilPapers and, derivatively, the Diversity Reading List in Philosophy as groupings of convenience that do not necessarily capture actual relations between topics. For example, we do not see why metaphysics and epistemology should constitute a single topic category or contain philosophy of religion or language. We propose to skip those groupings, moving directly to a somewhat more coarse-grained yet manageable list of ten items, including Aesthetics, Epistemology, Logic, Metaphilosophy, Metaphysics, Moral Philosophy, Philosophy of Language, Philosophy of Mind, Political Philosophy, and Social Philosophy. (We recognize that we have used traditionally Western names for our proposed topics. However, as with language and calendar systems, this choice of presentation is largely unavoidable, given the prevalence of these topic names in most, if not all, contemporary philosophical contexts.)
Expanding a category would reveal sublevel categories—for example, if Social Philosophy were to be expanded, topics such as the following would be presented: Philosophy of Class, Philosophy of Culture, Philosophy of Disability, Philosophy of Education. And similarly, if Logic were to be expanded, then topics such as the following would be presented: Formal Logic, Informal Logic, Mathematical Logic, Scientific Logic, and so on.
Using a broader range of categories prevents groupings in which seemingly ‘fringe and unimportant' topics and ‘core or important' ones are lumped together. Arranging categories in alphabetical order at all levels avoids any prioritization based on popularity, centrality, complexity, and so on. Excluding Miscellaneous and crypto-miscellaneous categories such as Philosophical Traditions ensures that marginalized content cannot be ‘hidden away’ within them. And lastly, in addition to replacing the History of Western Philosophy main category with a filter for time (discussed below), we will also provide a more neutral filter set, figures (discussed below), which will include philosophers who have had biographical works written about them.
From a technical perspective, the proposed model can easily be created by slight modifications to the index developed by PhilPapers: (1) ordering all categories alphabetically is a simple setting; (2) most entries classified under History of Western Philosophy, Philosophical Traditions, or Miscellaneous, will (and all of them should) also be classified under other categories, so removing these categories should not prevent users from searching for and finding their contents; (3) content classified in categories such as Kant: Metaphysics can be reclassified in bulk under Kant in figures and Metaphysics in topics. While this would require some work, it is also easily automated.
3.2 Time (When?)
Apart from searching through topic filters, users are also likely to want to look for works written in or about a specific time-period. This need is currently served in PhilPapers and the Diversity Reading List in Philosophy indexes by the chronologically arranged History of Western Philosophy, and parts of the much less structured Philosophical Traditions. Thus, a person looking for work on Ancient Greek philosophy of language would navigate to History of Western Philosophy and find, say, a category Plato: Philosophy of Language.
This approach has some serious disadvantages. Firstly, it is unparsimonious, as all texts found in the category Plato: Philosophy of Language should also be found in the topic category Philosophy of Language. Secondly, it perpetuates a division (‘the West and the rest’). Thirdly, even if ‘Western’ were dropped from the category name, the existing periodization (such as ancient, medieval, 17th century, 18th century) is only relevant to the Western tradition and does not map onto other traditions. For instance, is the Islamic Golden Age of philosophy Renaissance philosophy, medieval philosophy, or neither? It seems to depend on whose perspective we take. Finally, it effectively prevents cross-cultural comparisons, not allowing users to see easily what was happening in philosophy of language at a given time in different traditions.
We propose that instead of being included within topic categories, time should also be used as a filter, not unlike the filters found on shopping websites that allow users to select a price range of products to display. As illustrated in figure 4, users could simply type in a date range or use a slider feature, with a start date of (pre-fifth-century BCE) and end with (present).
Time filter with slider feature
On this approach, a user looking for ancient Greek philosophy of language would first select the category Philosophy of Language and then select a date range, such as fifth century BCE—first century. CE.
Technically, this could be achieved by assigning a date range to each database entry, based on the author's lifespan (for example, works by Anne Conway would be assigned the dates of 1631–1679, corresponding to her lifespan). If the user's search range overlaps with the author's lifespan, the index will display that author's works. Further, works by historians of philosophy could be assigned two (or more) date ranges: one for the time when the historian wrote their article or book (say, 2020), and a range for the historical individual's lifespan which their article or book covers (say, 1631–1679 for a book on Anne Conway). These works would then be found if a user's search range overlaps with either of those ranges.
Three worries might seem relevant here but are easily overcome. Firstly, the birth and death dates of some philosophers might be unknown. This is easily resolved by simply adopting the widest likely range. Secondly, the dates are based on the Gregorian calendar, which is not universally used. Here, we argue for a pragmatic solution similar to the one we adopted concerning a database's language: leave it open to the database, given its aims and audience. For the Diversity Reading List in Philosophy, though, we would adopt the Gregorian calendar as it is currently most universally used and admitting multiple alternatives might lead to too much confusion. Moreover, many philosophers and educators across the globe, already abide by this system. (Again, this choice is an uncomfortable one, as there are non-Western dating and calendar systems that will consequently be excluded, among them the Hijri, Ethiopian, Persian, Alexandrian, and Balinese Saka calendar systems.) Thirdly, users searching for, say, the works of Anne Conway might not know when she lived and thus not know what dates to input. However, this is easily solved (and contributes to the user's education) by simply searching for Anne Conway using any search engine, or by browsing figures (discussed below), and finding the relevant dates.
One might also worry about practical difficulties. Ensuring that entries are in fact tagged with the relevant date ranges will require alterations, and for larger databases such as PhilPapers, these alterations might require substantial effort. However, such alterations are also very easily automated to produce at least approximate results. Firstly, all entries already have the publication date assigned to them, and this field can simply be reused for the present purpose. Secondly, most historical entries that may have modern rather than historical publication dates attached, and modern entries focused on historical content will be likely classified under a relevant category within History of Western Philosophy or Philosophical Traditions, from which at least approximate date ranges can be automatically derived. For example, a 1994 article on Hume's Treatise could be automatically assigned its publication year as its first date range (alongside every other index entry), and then automatically assigned a second range of 1700–1800, which would roughly capture Hume's lifespan (alongside all other entries currently classified under 18th-Century Philosophy). While imperfect, such a fix would afford approximations at least as good as the ones offered at present, and benefit from being very easy to automate. Finally, there will be some entries not currently historically classified, thus not allowing for automatic date range assignment. However, in their case, no information is lost in the introduction of our proposed solution—and they should probably be classified anyway.
Naturally, achieving neutrality requires a certain trade-off: despite the fact that our proposed system is very easy to use, it is harder than the present one just by virtue of being different. People are used to using named periods (such as ancient, medieval, early modern, and modern). But although some philosophers might be more comfortable with this sort of periodization or find it easier for navigating philosophical history, using it presupposes that all history corresponds to the course of Western history. There are substantive debates as to how we might delineate such named periods within a Western-centric view of history (Shepard and Walker (Reference Shepard and Walker2009); Le Goff (Reference Le Goff2015)). But there are also substantive debates about taking a Western-centric view of history in the first place (Blankinship (Reference Blankinship1991); Bentley (Reference Bentley1996); Bin Wong (Reference Bin Wong1997)).
3.3 Traditions: Geographical (Where?) and Methodological (How?)
We also envisage traditions to be structured as filters. Instead of being a range, however, it would be a multiple-selection tool. A shopping website metaphor is once again useful: just as when shopping for clothes users can filter by specific brands, so a user who wants only to survey philosophical work that focuses on or writes from a North African or Marxist perspective could simply select a relevant filter. The selection of traditions should be inclusive and avoid existing biases treating some traditions as more important or valuable, and their list should be arranged alphabetically.
Arriving at the specific structure requires deciding which traditions should be included. Here, the need to balance accuracy and inclusivity with usability is particularly pertinent. A fine-grained list of traditions might be most fair and inclusive, but a filter forcing users to read and select a multitude of options will not make things easy, and having to supply those options will be time consuming for editors. Accordingly, we think that the best strategy is to rely upon two sorts of filters for tradition-types: Geographical Traditions and Methodological Traditions.
In Geographical Traditions, we propose the following list of categories: Central Asian, East Asian, European, North African, North American, Oceanian, South American, South Asian, Southeast Asian, Sub-Saharan African, and Western Asian Philosophical Traditions. Should more fine-grained distinctions be needed in a specific resource, the list could be nested with subcategories. For example, after selecting Southeast Asian Philosophical Traditions, users could select Abhidharma Philosophy, Confucian Philosophy, and Zen Buddhist Philosophy. Naturally, geographical divisions might not be relevant to some works, especially contemporary ones, but such entries can typically be classified with respect to methodology. For example, Shen-yi Liao (Reference Liao2021) has an excellent discussion of the difficulties associated with sorting individuals into traditions, geographically speaking. He is Taiwanese, but does this make him a Taiwanese philosopher? Or an East Asian philosopher, according to our suggested system? Largely, we think that authors (when possible) should situate themselves, and where they do not or cannot, contributors and editors should be as sensitive as possible to the circumstances of an author. We take this point to apply to more globalized traditions as well.
For Methodological Traditions, we would include as subcategories Analytic, Comparative/Intercultural, Critical (such as Critical Race, Feminist, Marxist, Postcolonialist, and Postmodernist), Cultural (such as African, Caribbean, Hebraic, Indigenous, and Persian), Hermeneutical/Phenomenological, Historical, Pragmatic, and Religious (such as Buddhist, Christian, Hindu, Islamic, and Judaic) Methodologies. We think that distinguishing Religious Methodologies from Cultural Methodologies as traditions is important. Including the latter is important because some methodologies are based primarily in specific cultures. And the former is important because, although some religious identities can overlap with certain cultural identities, there should still be options for those who want to distinguish between the two. For instance, someone may identify as Jewish in the sense of belonging to and operating within a Hebraic cultural framework but not identify as Jewish in the sense of being a practitioner of Judaism. As before, a more detailed and nested list of subcategories can be used where needed.
We think that the specific lists of traditions used might be adjusted as required. For example, a resource such as the Diversity Reading List in Philosophy might offer a more coarse-grained approach, thus nudging users to find resources they might have otherwise filtered out, while a resource such as PhilPapers might prefer a more fine-grained approach.
There are three points to note, though. Firstly, presenting these traditions in alphabetical order is best—but the geographical traditions seem best grouped together, and the methodological traditions correspondingly. Secondly, it would be acceptable for a given text to remain uncategorized under either Geographical Traditions or Methodological Traditions (although we estimate that this would be a rare occurrence) if it simply did not fit under any framework previously mentioned, or if it could not be identified as belonging to any specific geographical region. Instead of categorizing such entries under a Miscellaneous section, we think that topics, time, or figures would be sufficient to categorize the entry in such a way as to be accessible by users. And lastly, we also recognize that the methodological traditions that we have presented are slightly skewed toward more contemporary and Western ones. However, this aspect of the selection is incidental to historical circumstances wherein many methods have had an explicitly religious or cultural frame of reference while others have not. Still, for reasons of excessive space, time, and energy, it would be far too difficult for any philosophical database to identify and countenance all such religious or cultural frames—hence, they are unfortunately lumped together under Religious or Cultural Methodologies, and we rely on time, topics, geographical traditions, and figures to help identify more specific ones. Regarding the other methodologies, they have not been tied explicitly to religious or cultural frames of reference (although they might implicitly be), and they are geographically wide-spread and hard to localize. So, as a result, they are incidentally more visible, so to speak, in our selection.
Arriving at the specific shape of a traditions filter is difficult. We do not pretend to hold great authority here and would highly encourage a public discussion. However, we also think that there might be no great need to arrive at a single correct system. Instead, different resources could resolve this according to their specific needs. Should a need for a single system arise, a more fine-grained and nested subcategory division could be used, and specific resources could then select how many levels of granularity to use, matching their preferred trade-off between accuracy, inclusivity, and usability.
One could worry that adjusting existing databases such as the Diversity Reading List in Philosophy or PhilPapers to the new system would involve an inordinate amount of work. However, this job is once again easily automated, with the Traditions derived from existing categories. For example, texts currently classified under Confucian Ethics could be assigned the Religious or Cultural Methodologies (or both) in the methodological tradition filter and the East Asian geographical tradition. Given the existing analytic focus of the PhilPapers database, entries not currently classified under any Philosophical Tradition category, could be assigned as Analytic Methodologies, while those found in the History of Western Philosophy category, could be assigned the European geographical tradition. Such a solution would be incomplete and produce a somewhat messy result. However, the present index is also rather messy in this respect, and, as Fokt has argued (2020: 231), a somewhat work-intensive solution is needed regardless. We believe that if such work is to be done, it should be put into creating the most fair, inclusive, and usable system. Overall, we submit that our proposal does a better job of meeting these criteria.
3.4 Figures (Who?)
We envision figures as a filter as well, but one which is tradition- and time-neutral, with many philosophers who can usually be found under History of Western Philosophy or Philosophical Traditions, listed together alphabetically. In addition to those philosophers, and in alignment with the changing nature of the discipline, we also anticipate that with certain criteria for inclusion, figures will not just amount to a list of dominant, white, male figures such as Aristotle, Hume, Kant, Mill, and Plato. We leave it to different databases to cater to their own aims and audiences, with the appropriate labelings. But, in general and for the Diversity Reading List in Philosophy, we would suggest the inclusion of all philosophers who have had biographies written about them—which, to date and continuing into the future, appears to be growing into an impressive and diverse range of individuals. Among others, this criterion would include all of the philosophers from Rebecca Buxton and Lisa Whiting's (Reference Buxton and Whiting2020) The Philosopher Queens, which ranges from Ban Zhao and Diotima to Angela Davis and Azizah Y. al-Hibri. And, equally, it would also include such commonly known philosophers as Plato, Cicero, Confucius, Descartes, Leibniz, Kierkegaard, Russell, and so on. We recognize, of course, that some philosophical traditions do not have figures that are central to them. But, again, classifying a work under Figures is not mandatory, as we accept that this filter might be inapplicable to certain works. For such works, we anticipate that being classified under a geographical tradition or a methodological tradition would be sufficient (but even then, not mandatory). Practically, to ensure ease of use, the filter would be subdivided alphabetically: #, A, B, C, … Z.
The category # would stand for and allow for non- or semi-Latinate names/scripts if Latinized renderings are not preferred—so, names such as ᏍᏏᏉᏯ (Sequoyah). We could classify him in both # and S. But where C is expanded, what will follow are the further alphabetically arranged names of philosophers, for example, Cabral, Amílcar; Camus, Albert; Cavendish, Margaret; Cicero; Confucius; Conway, Anne. (Admittedly, this choice is also a slightly uncomfortable one, as there are other non-Latin alphabets or character systems. However, including all of those alphabets and systems would significantly increase the complexity and decrease the usability of the index. So we suggest that the alphabet or system used should simply follow the main language of the index in which it appears.)
In general, for a filter such as this, it is important to strike a balance between alphabetical categories containing so many names that it will overwhelm users and too few names that underrepresented individuals will be excluded due to restrictive criteria. What we have suggested seems to be more inclusive and fair in its representation than most other databases would be.
3.5 A User Experience
Imagine doing exploratory research for a paper on the topic of aesthetic experience. First, you find the category Aesthetic Experience under Aesthetics. Here you would find a list of works written on that topic at all times and in all traditions. While you might have approached this research looking only for modern analytic works, you would be immediately exposed to content you might not have otherwise looked for due to availability bias or lack of expertise in other traditions. At that point, you might narrow your search by selecting, say, 1950–2022 in Time, Analytic Methodologies in Methodological Traditions, or even Murdoch, Iris in Figures. These are conscious choices. But perhaps, you might instead think, ‘I never knew there were so many works on aesthetic experience written in East Asian traditions’, and decide to explore them. You might discover that current topics were also raised by Western Asian thinkers in the twelfth century CE. You might find perspectives that will shed new light and open new avenues for your research.
The existing indexing system, which almost exclusively lists modern analytic works in the relevant category, does not encourage similar reflections. Instead, it requires the user to find works written in East Asian or Western Asian traditions in separate, distant categories. In practice, its default approach is exclusivity, which requires users to opt in to seeing non-Anglo or non-analytic works, while in our system the default is inclusivity, requiring users to opt out of seeing works in all traditions or times.
4. Discussion: A New Normal?
How does our suggested indexing system compare to other philosophical databases such as PhilPapers? More specifically, to what extent, does the indexing system that we suggest recast what is normal within philosophy? To answer, we consider these systems in light of Fokt's (2020) primary concerns from our introduction regarding how PhilPapers has provided a particular sense of what is normal within philosophy as a whole.
Firstly, our system does not prioritize Western analytic philosophy. As outlined above, selecting any topic category from topics will, upon searching, display all works from all traditions and times in an alphabetical order. So, for example, Akan texts will be presented next to Confucian texts, analytic texts, and so on. Users will then be able to filter them by time, traditions, or figures to match their needs, but to do so they will need to opt out of seeing some content rather than simply not see it because of how the index is structured.
Secondly, our system displays everything in a value-neutral, alphabetical order, thus avoiding implicitly or explicitly implying that some topics are more important than others. Some value-laden choices remain, such as deciding on the selection of top-level categories or traditions. However, some selection is inevitable for pragmatic reasons. We believe that the system we proposed is more inclusive and maximizes the coverage of all possible philosophical topics, but we invite discussion.
Thirdly, we have done away with miscellaneous top- and sub-level categories, as well as crypto-miscellaneous categories such as Philosophical Traditions. Most, if not all, philosophical works are to some extent miscellaneous in content and can be easily classified under other categories where they can be found more easily. Simply doing away with them addresses a number of issues discussed above.
Fourthly, replacing the History of Western Philosophy category with time, traditions, and figures filters is more neutral and avoids presenting philosophy in terms of Western historical periodization or giving more weight to work that focuses on contemporary over historical philosophical questions and individuals. Should a database want to highlight modern or historical works, it could do so by preselecting a default date range, tradition, or figure for the users. In this regard, our system offers both adequate flexibility when needed, and ensures that time, traditions, and figures are noticeable and general features for most entries rather than only those labeled historical.
Finally, introducing the traditions filters ensures that our system does not prioritize any tradition, or present any tradition as the philosophy while simultaneously marginalizing others. By extension, it does not marginalize topics or people working in those traditions. Users can still filter out traditions they are not interested in, but the sole act of having to do so helps ensure that they do not wrongly conflate philosophy with analytic philosophy or any other tradition. Philosophy encompasses a wide variety of traditions and the proposed system represents it as it is—a multicultural, multi-geographical, multi-methodological, multi-perspectival, and multireligious community.
As our system addresses all of the issues previously identified, we propose it as better matched to meet the needs of the changing philosophical community. We believe it is fairer in its presentation of philosophical works because various marginalizing, paternalistic, and exotifying tendencies are rooted out. It is also more in alignment with how philosophers think their discipline should be: inclusive.
5. Conclusion
To what extent, and how, is it practically implementable for various philosophical databases, such as the Diversity Reading List in Philosophy or PhilPapers? We suggested ways that the practical difficulties involved can be addressed and offered solutions that can be implemented at a relatively low cost by virtue of being easily automated. There are good grounds for implementing the system that we have developed regardless. The discipline's current indexing practices are an artifact of a bygone era when it was normal to be much more exclusive—or, said another way, inclusion was not understood as important. As the discipline is changing, so are its practitioners’ conceptions of what it is or should be. Accordingly, it is right that the discipline's indexing practices follow suit and recast and broaden what is normal within them.
The general features of our proposed new indexing system recast historical and figure- or tradition-focused categories as filters, remove Miscellaneous categories, and implement alphabetical sorting. These modifications offer not only immense improvements in terms of fairness and inclusivity but also user efficiency.
We further propose some specific solutions, such as the exact composition of top-level topic category filters, a list of geographical and methodological traditions, and specific criteria for highlighting certain figures in the hope that our proposal will inspire a public discussion aimed at optimizing them.
Our next step is to implement our system within the Diversity Reading List database. As mentioned, the mission of the database is to inspire and facilitate positive change while also reducing the amount of work and time required to achieve it. We hope to do just that: create a working proof-of-concept indexing system that will become the first step in the right direction and the start of a much broader discussion, subsequently facilitating near-future practical changes across the discipline.
Open access
