1. Introduction

To answer the question why some of Shakespeare’s sonnets appear more often in anthologies than others, Thomas Kullmann in his article “Anthologizing Shakespeare’s Sonnets” proposes that, instead of focusing on ideological or purely aesthetic reasons—like a society’s views on gender or “literary eminence” (65) in terms of composition and imagery—, one should look at the different communicative situations the dramatis personae of the sonnets are placed in, and the predominant function of language in each sonnet.

Kullmann’s approach adds yet another way of dividing the sonnets into meaningful groups to the various models suggested in the past. One of the oldest and most enduring of these was undertaken by Edmond Malone in his 1780 edition of the Sonnets when he suggested that we should organize the poems according to addressee, with the first 126 sonnets being directed to a male youth while the remaining ones refer to a mistress (see 579). Other classifications focus on the sequence’s plot (cf. Pequigney 1985, Rudenstine 2015), or its lack thereof (Burrow 2002, Schiffer 2007, Edmondson and Wells 2013, Weidle 2025), recurring keywords (Monte 2021), numerological patterns (Duncan-Jones 2010, Booth 2000), themes, as, for example, jealousy (Pequigney 1985) or proximity to death (Cousins 2011), character constellations and configurations (rival poets- or love-triangle sonnets, cf. Paterson 2010 and Duncan-Jones 2010), and so forth.

In the following, I will first briefly summarize Kullmann’s model, then address what I believe to be some of its problems, and finally suggest expanding his focus on communicative modes by also considering the sonnets’ thematic, aesthetic, formal, and rhetorical qualities to explain why some of them have been more popular than others.

2. Communicative Modes

To state his case, Kullmann begins by focusing on sonnet 2 (“When forty winters shall besiege thy brow”), a sonnet that, although very [→ 291] popular in the seventeenth and eighteenth centuries, is included in only three of the altogether 33 anthologies published after 1861 that Kullmann takes into account. Why, he asks, has this sonnet fallen out of favour although it indubitably possesses literary merits? It presents a “clear-cut argument,” abounds in “original images, organized in coherent conceits or image clusters,” is “rich in imagery,” provides “an intricate and unusual point of view” (70), and is characterized by “formal perfection” (71). The answer, according to Kullmann, lies in its mode of communication. Comparing sonnet 2 with the often-anthologized sonnet 30 (“When to the sessions of sweet silent thought”), he notices that the speaker, who in the former sonnet does not refer to himself at all and whose “messages are [all] second-person messages” (72), in sonnet 30 “provides a chain of first-person messages” (72). Similarly, when comparing sonnet 2 to sonnet 116 (“Let me not to the marriage of true minds”), Kullmann ascertains that the latter does not contain many first-person messages, that the “message is more abstract” than in sonnet 2 and that it “does not involve an addressee” (72). So, while these sonnets “reach the highest standards of poetic excellence” (72), they each foreground different functions of language as identified by Roman Jakobson in his seminal essay “Linguistics and Poetics” from 1960. Kullmann concludes “it is obvious that in sonnet 2 the conative function predominates, as opposed to the emotive function in sonnet 30, and the referential function in sonnet 116” (73). In applying Jakobson’s model to all the sonnets in Shakespeare’s sequence, Kullmann identifies 26 sonnets with a predominant emotive mode, 33 sonnets which privilege conative statements, 22 sonnets in which the referential function predominates, 14 sonnets that are self-referential (Jakobson’s metalingual function), nine sonnets dealing with the speaker’s triangular love relationships that “cannot easily be classified as either predominantly conative or referential” (74), and 50 sonnets with a combined “I-and-thou mutuality” (89). In this last group, the emotive and conative functions are “subtly intertwined with one another” (75). In fact, Kullmann argues that this interweaving of first- and second-person messages is a [→ 292] general feature of the Sonnets and that the conative function is rarely completely absent from the poems.

To substantiate and correlate his hypotheses about the preference of emotive and referential over conative sonnets with empirical data, Kullmann went through an impressive list of 38 anthologies and selections of sonnets that were published between 1783 and 2023. In the 45 sonnets that were anthologized between five and nine times, and in the 27 sonnets included ten or more times, Kullmann notices a “strong bias in favour of first-person message and non-personal pronouns” (77). According to Kullmann, this practice of including sonnets with first-person and non-personal messages began in the second half of the nineteenth century with the collection by Palgrave (1861). Of the 20 sonnets included in this anthology, more than half focus on first-person messages and “non-personal wisdom” (78). For Kullmann, Palgrave’s predilection for emotive sonnets corresponds to the strong influence Romantic notions of poetry still held over the Victorians, i.e. the belief that poetry should express the speaker’s “suffering caused by the ways of the world and the human condition” (78). Referring to collections published in Poland, Germany, and India, he suggests that the (basic) tendency towards anthologizing first-person sonnets has continued up to the present and “transcended national boundaries” (79). New trends, however, have emerged in some of the more recent anthologies. For example, in the Bedford Anthology of World Literature (2004) Kullmann identifies a decrease in first-person sonnets and a heightened focus on self-referential and non-personal sonnets while he also states that in some of the more recent editions we find a reduction of I-and-thou sonnets.

3. Some Problems

As promising as Kullmann’s focus on the communicative modes in the sonnets may be, his argument that from mid-nineteenth century onwards anthologies have shown an increased preference for sonnets with an emotive, referential, and metalingual focus, while conative as [→ 293] well as combined emotive and conative sonnets have largely been neglected, does not fully convince me. There are, I believe, some inconsistencies and incongruities in his argument that warrant closer inspection.

First, I am not too sure whether the empirical data Kullmann provides does represent enough evidence to substantiate the main argument. He finds that, of the 45 sonnets anthologized five times or more, thirteen are first-person messages and seven express non-personal reasoning. According to him, this is indicative of a “strong bias in favour of first-person message and non-personal poems” (77). This, however, seems to be reading too much into these numbers. I am also not quite convinced that the fact that merely 14 of the 27 sonnets listed in ten or more anthologies are first-person or non-personal messages is sufficient evidence to corroborate Kullmann’s claim.

Apart from the equivocal validity of the data provided, I also see a few problems in the methodological approach. While it may prove productive to classify the sonnets according to their mode of communication based on Jakobson’s model (emotive, conative, referential, metalingual, etc. functions), I have some doubts about the final categories Kullmann comes up with, since they combine Jakobson’s functions of language with grammatical and thematic criteria. I wonder, for example, how productive it is to subsume the sonnets referring to someone “in a third person,” to abstract concepts, and to “a mythological story” (74) under the referential category. Similarly, how meaningful is it to create separate categories for the sonnets dealing with “triangular relationships” (89) and “I-and-thou mutuality” (74), and how do these groups relate to Jakobson’s functions? I also think that, if one applies Jakobson’s model to the analysis of the sonnets, one needs to differentiate between those poems with a dominant metalingual function and those with a prevalent poetic function since there is a difference between sonnets thematizing the code (language and writing) and those thematizing the message itself (poetry). Compare for example the “Sonnet letters” (Edmondson and Wells, All the Sonnets 28) 26 (“Lord of my love, to whom in vassalage”) and 77 (“Thy glass will show thee how thy [→ 294] beauties wear”) or the notebook sonnet 122 (“Thy gift, thy tables, are within my brain”), all three of which thematize language and writing in more general terms, with those sonnets on the writing of poetry which are therefore much more “self-referential” (Kullmann 74; cf. the sonnets that I grouped in the theme cluster “writing” in Weidle, ch. 5).

At times, Kullmann himself seems to realize that his mode-based model may not be able to fully explain the popularity of specific sonnets throughout the ages. While arguing that sonnets with a predominantly emotive and referential function are preferred over others, he at the same time somewhat paradoxically concedes that there are sonnets that appeal to the reader because of their thematic concerns: sonnets 73 (“That time of year thou mayst in me behold”) and 116, for example, have remained popular with teachers because of “their unexceptionable messages about ageing and true love, and avoidance of the issues of sexuality and the young man’s beauty” (69), and sonnets 20 (“A woman’s face with nature’s own hand painted”) and 129 (“Th’expense of spirit in a waste of shame”) have attracted heightened interest precisely “for the opposite reason: ‘Sex sells’” (69). In the end, Kullmann seems to undercut his own argument and findings about the impact of communicative modes on a sonnet’s canonicity when he says that “[c]hanges in the canon are rather due to the impact of cultural movements like Romanticism or a more recent interest in iconoclasm and non-heteronormative sexuality” (83).

4. A More Comprehensive Model

To arrive at what I believe to be a more sustainable model, I suggest to use Kullmann’s “List of Sonnets Found in Anthologies” (84-88) and focus only on those sonnets that have been included considerably more frequently in collections than the rest. This may allow us to expand Kullmann’s focus on communicative modes to also include thematic, formal, stylistic, and structural features to explain a sonnet’s popularity. [→ 295]

On Kullmann’s list there are 19 sonnets that stand out from the others since they appear in at least 15 collections: sonnets 18, 29, 30, 33, 55, 60, 64, 71, 73, 94, 97, 98, 104, 106, 116, 129, 130, 138, 146. Of these sonnets, almost one-third belongs to Kullmann’s category of “non-personal reasoning” (89) on abstract concepts, such as sonnet 60 on “maturity and decay” (Duncan-Jones 230), sonnet 64 on “the operation of time” (238), sonnet 94 on beauty’s “obligation to behave virtuously” (298), sonnet 116 on “true love” (342), sonnet 129 “[o]n lust” (Edmondson and Wells, All the Sonnets 29), and sonnet 146 on the “poet’s soul” (28). Other sonnets, such as numbers 18, 55, 106, and 130, reflect on the functions and limitations of poetry, and sonnets 30 and 73 are characterized by a high degree of formal and rhetorical sophistication, the former exhibiting the “exactness of Shakespeare’s psychological portraiture” (Vendler 167) and featuring a dense phonetic structure with multiple instances of polyptoton and “moany Os and sighing sibilants” (Paterson 91), and the latter being widely regarded as one of the most accomplished sonnets of the sequence (see Paterson 210). In other sonnets from this group, the speaker communicates experiences and emotional states readers can readily relate to, as in sonnets 97 and 98 on being separated from a beloved person, in sonnet 104 on unwavering affection in the face of time and age, or in sonnet 138 on “mutually dependent self-deception” (Duncan-Jones 390).

This means that most of these 19 sonnets that have appeared most frequently in anthologies since George Kearsley’s The Beauties of Shakespeare from 1783, the earliest anthology consulted by Kullmann, are on abstract concepts, poetry, and/or relatable states and experiences, and/or they are characterized by a high degree of formal and stylistic sophistication. Notwithstanding the fact that some of these qualities are also connected to a referential and emotive communicative mode, as pointed out by Kullmann, I believe that it is above all these qualities and thematic concerns that have been responsible for these sonnets’ ongoing appeal.

What most of these often-quoted poems also have in common, is that they are “stand alone sonnets” (Edmondson and Wells, Shakespeare’s [→ 296] Sonnets 33), which is to say that they do not need the context of neighbouring poems or additional context for readers to fully understand them. While sonnet 45, for example, a poem that is not included in any of the 38 anthologies consulted by Kullmann (see “Appendix I”), continues an argument on the four elements from the previous sonnet (“The other two, slight air, and purging fire,” 45.1), sonnets 116 on constant love and 129 on lust, although being very different in tone and outlook, are self-sufficient, impersonal, and abstract reflections on human experiences and desires that strike a chord with most readers because, as Stephen Booth remarks in his commentary on sonnet 116,

abstract general assertions do not feel any truer than their readers already believe them to be […]. The attraction of abstract generalizations is the capacity they offer us to be certain o’er incertainty (115.11), to fix on a truth that allows for and cannot be modified by further consideration of experience or change in our angle of vision. (387)

I have argued elsewhere that the long-lasting appeal of Shakespeare’s Sonnets derives from a combination of formal, thematic, and structural features that together create their “passionate rationality” (Burrow 91). The sonnets, to varying degrees and due to their “antithetical form” (Tetzeli von Rosador 578; my translation), are characterized by logical reasoning: they develop their argument by means of a syllogistic structure with two propositions followed by a conclusion (see also Weidle 17-18). Often, however, this conclusion turns out to be flawed or self-defeating, which has to do with the other, passionate side of the collection: the sonnets deal with a range of intense and contradictory emotions, states, and phenomena, such as love, sex, hate, guilt, blame, jealousy, rivalry, friendship etc. In many of the sonnets grouped by Kullmann in the emotive or referential group (and thus belonging to the more popular sonnets) the speaker’s struggles in rationalizing his conflicted desires and feelings are more pronounced than in the other sonnets.

Kullmann’s application of communicative modes to the Sonnets is innovative and promising. I find, however, the empirical evidence provided so far not fully convincing. Moreover, I suggest that instead of [→ 297] relying solely on communicative modes to explain a sonnet’s popularity, one should also take into account qualities such as abstractness, self-reflexivity, relatability, stylistic sophistication, stand-aloneness, and a high degree of “passionate rationality,” all of which we find in many of the most often-quoted sonnets.