Doing philosophy as opening parentheses: quantifying the use of parentheses in Stanley Cavell's style

ABSTRACT

The aim of this paper is to say something significant about Stanley Cavell's style. To accomplish this task, we adopt a distant reading approach, quantifying what seems to be an idiosyncratic use of parentheses. After outlining our methodological approach and the choices of texts from Cavell's corpus, we will present the results of our quantitative analysis. Two kinds of results will be presented and interpreted: the result of a comparison between Cavell and other authors (i.e. why Cavell's use of parentheses is exceptional) and the result of a quantitative analysis of Cavell's texts (i.e. how Cavell uses parentheses throughout his books). For both results, we will provide our own interpretation. In the conclusion, we will draw the moral of this parenthetical story, hoping to open future parentheses.

KEYWORDS:

• Stanley Cavell
• distant reading
• style
• metaphilosophy

Acknowledgments

We would like to thank Guido Bonino, Paolo Tripodi, Marina Antonia D'Amico, Marco Fornaseri, Daniel Gallano and Carlo Debernardi for the help and the useful insight. We also would like to thank Isaijah Johnson, Michael Kane, Ryan Lam and Alice Re for their feedbacks.

Disclosure statement

No potential competing interest was reported by the authors.

Notes

1 Of course, we have in mind the approach outlined by Franco Moretti in Distant Reading (2013), which is an approach that can also be adopted without the help of the computer. Moretti was our primary source of inspiration, therefore we will use the label ‘distant reading’. Nevertheless, we have in mind various types of similar approaches that fall within the wider label of ‘digital humanities’ and, generally speaking, quantitative methods in humanities. Here you find a list of people who have already done such works in the history of philosophy and in the history of ideas: Betti and van den Berg (2014), Betti and van den Berg (2016), Betti et al. (2019), De Bolla et al. (2020), Bonino and Tripodi (2019), Bonino, Maffezioli, and Tripodi (2020), Malaterre, Chartier, and Pulizzotto (2019), Petrovich (2018) and Petrovich and Buonomo (2018). In this context, it is worth reminding that both stylistics and digital humanities have long histories. For instance, in 1940s, Josephine Miles was counting adjectives in Romantic poets, and people like Robert Cluett were already computing sentence lengths and punctuation marks in Hemingway and other modern writers in the 1980s and 1990s. A key early text is also Burrows's Computation into Criticism (1987). Stylistics is also known as ‘authorship attribution’ in some contexts; this also has a long history overlapping with corpus and forensic linguistics. In the 1990s and 2000s, stylistics was famously critiqued by authors like Stanley Fish. We hope that our interpretations and data allow the present paper to disprove Fish's and others scepticism about these methods. We thank an anonymous reviewer for making us more aware of the richness of this history and of its various critics.

2 We can provide here a minimal bibliography of the studies that address, more or less directly, the question of style in the philosophy of Stanley Cavell. One of the them is certainly that of Timothy Gould: Hearing Things: Voice and Method in the Writing of Stanley Cavell (1998). Gould prefers to refer to the concepts of ‘voice’ and ‘method’, but he is well aware that ‘style’ represents a central question in Cavellian philosophy. And Peter De Bolla is well aware of that fact too when he decides to dedicate a study to a concept related to ‘style’, namely ‘tone’. See his ‘The Tone of Praise’ (2020). One of the more direct studies of the style in Cavell is then to be found in Áine Mahon. Indeed, in her book The Ironist and the Romantic: Reading Richard Rorty and Stanley Cavell (2015), Mahon devotes an entire chapter to a description of their philosophical styles (and the chapter is aptly titled Stylists of the Philosophical). Furthermore, it remains of interest the list provided by Michael Fischer in Stanley Cavell and Literary Skepticism (1989) of Cavell's detractors – who, criticizing his style, nevertheless underlined its centrality for a critical analysis of Cavellian philosophy. Many of the authors cited by Fischer refer to parentheses as one of the weapons of the much-criticized Cavellian ‘self-indulgence’. For a detailed answer to these criticisms of Cavell's style as ‘self-indulgent’ see Stephen Mulhall ‘On Refusing to Begin’ (2005). More recently, we can also refer to a forthcoming paper by one of the authors: ‘Lingering: Wittgenstein, Cavell, and the Problem of Style’ (forthcoming).

3 This choice can appear restrictive, leaving out of consideration some works from these authors which can be important from a stylistic point of view. Nevertheless, we are confident of our choice for two reasons. First, this is a paper on Cavell's work and that has to remain the focus of our exploration. The other authors are considered only to highlight the exceptionality of Cavell's style and not to discuss their own style. Second, the most cited opera of every author is probably the one that closely resembles the common idea of their style, simply because it is the more read and studied. To keep the comparison fair, we also consider only The Claim of Reason when comparing Cavell with the other authors.

4 The specific version considered for every book, including the chosen translation, is reported in the bibliography.

5 The pattern used for the regex is, in python dialect, for the sentences inside the parentheses r'\([ˆ()]*(?:\([ˆ()]*\)[ˆ()]*)*\)|\[[ˆ\[\]]*(?:\[[ˆ\[\]]*\][ˆ\[\] ]*)*\]'and r'[\.\?\! ]\s*(\( [ˆ()]*(?:\([ˆ()]*\)[ˆ()]*)*\)|\[[ˆ\[\]]*(?:\[[ˆ\[\]]*\][ˆ\[\]]*)*\])' for those after a mark.

6 See the source files available at https://github.com/TnTo/CavellParentheses (particularly the functions clean and extract_brackets in the libcavell.py file) for a comprehensive list of the automated text-processing performed.

7 See the function tokenize in the libcavell.py source file.

8 We want to highlight that this measure is substantially equivalent to measuring the number of tokens inside the parentheses that are a sentence on their own. But some corner cases in either direction make these two measures different, for example a parenthesis after an etc. or a dot at the end of a sentence lost during the digitalization. We opted for the measure simpler to be measured.

9 This statistical test tries to measure the likelihood of the extraction of k white ball in a sample of n balls from an urn in which only a fraction p is white. In our case, we want to measure how likely it is that a word occurs a given number of times in a sample (the words inside the parentheses) given the frequency in the population (the text). This is a paradigmatic situation in which the binomial test, which is defined between a sample and the population, is better suited than a non-parametric test between two populations (like the Mann–Whitney–Wilcoxon test).

10 A longer list of word (p-value < 0.1) is available with source code.

11 In his ‘Wittgenstein's Texts and Style’ (2016), David Stern makes a list of the devices typical of the later Wittgenstein: multiple voices, thought experiments, provocative examples, striking similes, rhetorical questions, irony, parody. Among them, Stern writes, the use of multiple voices is the most important aspect of Wittgenstein's style.

12 A more comprehensive list is reported in online appendix at https://github.com/TnTo/CavellParentheses/tree/main/paper/appendix.pdf. In The Senses of Walden, the tokens highlighted as more frequent inside the parentheses are god, he, man and i (which, looking to the unprocessed data, appears to be the roman number 1 and not the pronoun I). In The World Viewed nothing of interest stands out: in particular for the second part the only words listed are you (discussed in the text above) and marquis, a character from the film Rules of the Game.

13 See tables in online appendix at https://github.com/TnTo/CavellParentheses/tree/main/paper/appendix.pdf.

14 The complete list is reported in online appendix at https://github.com/TnTo/CavellParentheses/tree/main/paper/appendix.pdf.

15 See online appendix at https://github.com/TnTo/CavellParentheses/tree/main/paper/appendix.pdf.