Abstract
This article presents a corpus stylistic analysis of the embodiedness of Australian-born song writer Nick Cave’s (b. 1957) song lyrics. More specifically, the article explores the use of verbs in Cave’s similes and considers how such constructions, labelled dynamic similes, give rise to a sense of eventfulness anchored in embodiment. Through corpus methods, the prominence of verbs in Cave’s similes is first investigated in contrast to two sub-corpora of the British National Corpus, BNC Fiction & Verse and BNC Newspapers. The comparison shows that Cave’s similes are richer in this regard than the sub-corpora. Further, building on a semantic categorisation of the data, two salient types of simile are analysed more closely: similes containing verbs of movement and psychological processes. Applying insights on profiling verbs from Cognitive Grammar (Langacker, 2008), as well as the notion of eventfulness in lyric texts (Hühn, 2016), the article finds that the dynamicity of Cave’s similes typically serves to convey the intensity of mental processes, even with verbs of movement, and is frequently reinforced through progressive verb forms and accompanying sensory lexis. The article additionally concludes that by evoking experiential similarities (Dancygier, 2022), Cave’s similes often invite embodied engagement on the recipient’s part.
1. Introduction
Nick Cave is among the most well-regarded lyricists in the contemporary music scene, praised for his unique style of writing rich in cultural references and evocative scenes (see e.g. Eaton, 2022; Wray, 2021). However, despite a career spanning four decades, Cave’s lyricism has not garnered a lot of academic attention, and most of the existing commentary focuses on the overarching themes and cultural influence of his work. From a stylistic point of view, an interesting aspect that has been highlighted in this research is the embodied dimension of Cave’s lyrics (e.g. McCredden, 2009; Van Elferen, 2013), manifested for example in depictions of physical intimacy and confrontation. In particular, the combination of bodily and spiritual lexis has been mentioned as an engaging feature: McCredden (2009: 167) writes that ‘[t]he sacred […] and the profane […] are in dynamic and conflicting conjunction’ in Cave’s lyrics, linked with the often concurrent themes of desire and destruction (Barfield, 2013) and regularly shown through fictional characters’ immoral decisions (Eaglestone, 2009). Taking cues from these observations, the present article explores how such dynamicity is communicated to listeners linguistically. Through a combination of close reading and corpus methods, the aim is to offer new insights into Cave’s style as a lyricist.
Given that actions are integral to conveying the embodied dimension of various kinds of scenarios, we identify Cave’s use of verb phrases in similes as a fitting and sufficiently narrow focus for the investigation. As has been established in Cognitive Linguistics, bodily experience forms the foundation for many metaphorical language patterns (Gibbs, 2014), and simile is particularly suited to a corpus stylistic investigation of such patterns due to the linguistic explicitness of the figurative comparison, typically signalled with ‘like’ (e.g. Goatly, 2011). Furthermore, whereas the metaphorical use of verbs has been researched extensively, the role of the verb phrase in simile has not been discussed very much at all, particularly as regards poetic effects. We aim to demonstrate that attention to dynamic similes, that is similes containing a non-stative verb (phrase), elucidates how embodiment contributes to creating a sense of eventfulness in Cave’s lyrics. With the help of corpus methods, the article also offers quantitative insights on the frequency and type of simile used by Cave. Here, Cave’s lyrics are compared to two sub-corpora of the British National Corpus (henceforth BNC): BNC Fiction & Verse and BNC Newspapers, which provide samples of natural language from both (somewhat) similar and dissimilar genres.
While simile in poetry is a thoroughly explored topic (for an overview, see e.g. Brogan and Smith Richmond, 2016), stylistic research into song lyrics is quite scarce overall (West, 2019). Previous studies focusing on figurative language in song lyrics include Steen’s (2002) exploration of Bob Dylan’s ‘Hurricane’, Schröder’s (2012) corpus-based investigation of rap lyrics, Price and Wilson’s (2019) discussion of Tom Waits’ ‘Emotional Weather Report’ and Hidalgo-Downing and Filardo-Llamas’ (2020) multimodal analysis of three U2 songs. Computational methods have been applied to the analysis of song lyrics mostly in large-scale genre-focused research endeavours 1 (e.g. Bridle, 2018; Kreyer, 2015; Kreyer and Mukherjee, 2007; Werner, 2019, 2021). As a single-author study with a specific linguistic focus, the present article does not directly adopt methods from these investigations, however. Another methodological caveat involves the multimodality of the genre. While the article cannot delve into musical aspects as such, we will attend to a particular limitation of a zoomed-in stylistic analysis of song lyrics, as raised by Morini (2013: 289): ‘An exclusively textual analysis […] is both incomplete and slightly incongruous, presupposing as it does a “musical reader” who listens to the album attentively, with booklet or printed lyrics in hand.’ Similarly to Voice and Whiteley (2019), who extend Morini’s multimodal method to analysing attention in Outkast’s ‘Hey Ya!’, we argue that a cognitive approach to similative meaning can help mitigate the issue. This is discussed in more detail in section 2.2 and in the analysis.
Related to this, it is noteworthy that recent media coverage on Nick Cave frequently takes up Cave’s interaction with his fans on matters of art, emotional life and religion (e.g. Balstrup, 2020). Cave publishes a regular newsletter called The Red Hand Files in which he answers a wide range of questions sent in by fans. Thus it seems appropriate to incorporate a readerly angle into an investigation of Cave’s similes, in an effort to illuminate how his songs are ‘simultaneously discomforting and appealing’ (Van Elferen, 2013: 179). In a study on the use of similes in Brexit discourse, Dancygier (2021: 668) argues that ‘similative figuration provides an experientially rich and vivid frame to help model the potential emotional response to an object or event’. Our aim is to uncover the potential that Cave’s similes have as regards providing an experiential frame for responding to the scenarios portrayed in his songs. While Dancygier’s study builds on Blending Theory, we mainly employ Cognitive Grammar (Langacker, 2008): the qualitative analysis will focus on the kinds of processes that are profiled in the verb phrases in dynamic similes. To zoom in on possible embodied effects, the discussion will be limited to two prominent semantic categories of verbs that appear in Cave’s lyrics and connect to events, actions and perceptions.
The questions the article seeks to answer are: • How prominent is simile in Cave’s writing, compared to a sample of other fictional and non-fictional texts? • To what extent does dynamicity, as dynamic verbs, occur in Cave’s similes, compared to a sample of other fictional and non-fictional texts? • How does the verb phrase contribute to the realisation of embodied experiences in Cave’s similes?
To address these questions, quantitative corpus linguistic methods will be combined with close textual analysis. We begin with a brief review of research on simile and an account of how events may be understood in lyric texts. After a succinct method description, the article goes on to present the quantitative findings, followed by a qualitative discussion of selected similes that represent semantic categories prominently emerging from the quantitative analysis.
2. Theoretical background
2.1 Simile
Although there is no consensus among researchers about the definition of simile, particularly when it comes to its distinctness from metaphor, it is usually agreed upon that simile is an especially visible mode of figurative comparison (Dancygier, 2021; Moder, 2008; Sullivan, 2014). Within the cognitive paradigm, which tends to view simile as a linguistic realisation of metaphorical thought patterns, Sullivan argues that ‘simile is relatively brief, but evokes metaphor explicitly and visibly’ (2014: 350). Similarly, Moder (2008) makes the point that simile cues a relatively instantaneous mapping, as opposed to being extended in the discourse. Moder’s findings, drawn from a corpus of spoken news discourse, suggest that simile typically functions ‘to consciously draw the listeners’ attention to the upcoming mapping.’ (Moder, 2008: 318) Though the data giving rise to this conclusion is non-literary, the notion appears relevant to the present study as well, given that song lyrics are typically heard, not read.
In the literature on simile, two strands of inquiry are of particular interest to our investigation: studies with a communicative orientation and studies that elaborate on the linguistic make-up of similes beyond the prototypical X is like Y construction that is not very common in actual language use (Dancygier, 2022). Combining the two foci, Israel et al. argue that ‘similes serve the basic rhetorical functions of description and evaluation. This is largely a consequence of their form, and the fact that a simile necessarily features a comparison construction predicated of an explicit target.’ (2004: 130) Along the same lines, Goatly claims that simile ‘lacks the initial contradictoriness of metaphor’ (2011: 251) and that many similes ‘are relatively explicit about the Grounds of the metaphor’ (2011: 195). In Goatly’s approach, based on Relevance Theory, the similarities between the source (Vehicle) and the target (Topic) are referred to as the Grounds, where Israel et al. (2004) use the term tertium. For both Israel et al. (2004) and Goatly (2011), a relative ease of interpretation is characteristic of simile. In sum, this stems from similes (in contrast to metaphor) possessing the capacity to manipulate the recipient’s attention through the Grounds (Dancygier, 2022: 208; Tartakovsky et al., 2019: 187).
However, Goatly goes on to point out that certain interpretive challenges may occur if ‘[t]he Grounds stated in the verb phrase are […] only partial’, allowing for ‘other Grounds to be exploited or created.’ (Goatly, 2011: 195) In a poetic context such exploitation is of course to be expected. Here, attention to the complete verb phrase is crucial. For example, in the simile ‘my pyjamas clung to me like a shroud’, from Cave’s ‘Hallelujah’ (2001), the complement ‘to me’ underscores that being wrapped up in a shroud should be thought of as similarly subjective as the acute sensation of clingy pyjamas. This, in turn, connects to an overarching sense of threat experienced by the song’s protagonist, created through subtle references to death, as in the earlier ‘[m]y typewriter had turned mute as a tomb’. Noting such details may illuminate to what extent similative figuration actually is explicit in conveying a resemblance, and therefore descriptive or evaluative. Indeed, Israel et al. argue that ‘[s]imilarity [...] is largely a matter of construal—it all depends what one is focusing on.’ (2004: 126) We will return to this aspect of conceptualisation later on.
In a corpus study that also incorporates literary data, Wikberg (2008) sheds some light on the role of the verb phrase in simile. Of particular interest is the linguistic pattern verb + like a, which Wikberg (2008: 141) characterises as the most innovative of the linguistic patterns included in her study. As regards verbs, Wikberg finds only one clear pattern across the BNC data: the predicative subcategory, which consists of perception verbs, makes up 52% of all instances of V like a (Wikberg, 2008: 138). Wikberg offers the explanation that these verbs convey sensory experiences, which are hard or even impossible to communicate using other means. As such, these types of similes are arguably not especially figurative, even if they occur most frequently in the Imaginative domain of BNC, as indicated by Wikberg (2008: 139). Otherwise the data includes a range of verbs that take the like-adverbial of manner; Wikberg states that they ‘tend to express events, states and very specific types of behaviour, the following like-phrase adding further precision.’ (2008: 139) Wikberg (2008: 140) also mentions that this range is generally typical of Imaginative discourse.
Another observation Wikberg makes is that in the part of the simile following the V like a pattern, there is often a blend of metaphor and simile (Wikberg, 2008: 140). In his account, Goatly talks about ‘Metaphorizing Similes’, instances where the simile acts as ‘a framework for extra metaphors’ (Goatly, 2011: 196). This tendency is something many scholars have noted, including Israel et al. (2004: 130), and it demonstrates how there is obvious potential for extending the communicative reach of similes in creative contexts. Gargani argues that scholars often ‘account for such effects in terms of impressions which establish themselves […] over the course of a whole poem’ (2016: 56). However, by examining concordance lines in context, we aim to attend to the potential dual nature of simile as both conceptual and procedural, as discussed by Gargani based on Walaszewska’s Relevance Theoretic account (2013). This involves the receiver encoding an ‘ad hoc’ concept by expanding on the conceptual content conveyed by the part following ‘like’, which will ultimately ‘help the hearer identify relevant properties attributable to the subject.’ (Walaszewska, 2013: 331)
This procedurality may require considerable effort in poetic contexts, including song lyrics, even if their medium-specific properties differ somewhat from poetry (cf. Holm, 2025; West, 2019). In a corpus study contrasting standard and non-standard simile, Tartakovsky and Shen (2019) found that non-standard simile occurs surprisingly frequently in poetic contexts compared to non-poetic corpora. In another article, they present an analysis of the grammar of non-standard simile, pointing out that grammatical prioritisation of the target in the topic position is a violation of expectations and requires active efforts to process (Tartakovsky and Shen, 2018: 105). Tartakovsky and Shen further argue that non-standard poetic simile possesses a ‘hit-or-miss potentiality’, prompting the recipient to rethink the source or the ground (target) in order to bring the simile elements together (2018: 117–118). Fishelov (1993: 18) conceptualises this as a ‘tensional balance’ that involves grappling with deviation whilst drawing on recognisable similarities on a semantic and structural level. While Fishelov finds in a small-scale empirical study that interpreters tend to lean on familiar semantic categories even in the case of novel or non-standard similes (2007: 84), Moon (2011) draws attention to the fact that dissimilarity, for example by way of uncertainty, may be a crucial component in similative meaning, especially in literary contexts.
At this point, it seems pertinent to outline how simile is viewed in the present study. We follow the definition used by Wikberg (2008: 128), ‘[a] simile can be defined as a figurative expression used to make an explicit comparison of two unlike things by means of the prepositions like, (as) … as or the conjunctions as, as if, as though.’ As the scope of the present study is limited, only simile constructed with the preposition like is investigated. Of course, literal comparisons and similes can follow the same linguistic structure and the difficulties found in disambiguating the two are discussed by both Wikberg (2008) and Tartakovsky and Shen (2019: 209–210). While Wikberg (2008) differentiates simile from literal comparisons on the basis of reversibility, arguing that literal comparisons are reversible, Tartakovsky and Shen specify that similes compare two unlike things through a ‘marked semantic distance between the target terms or, put differently, target and source need to belong to different taxonomies’ (2019: 209). We find Tartakovsky and Shen’s expanded description of what is meant by ‘unlike’ helpful for disambiguation.
Therefore, simile is here understood according to the following criteria: (1) The explicit comparison marker like is used; (2) The comparison involves two unlike things, where the target and the source belong to different taxonomies.
2.2 Understanding events in lyric texts
Cave’s song lyrics are often described as narrative, even including third-person accounts of fictional characters’ actions, although by his own declaration (Cave and O’Hagan, 2022: 4–5) this tendency has waned in the later works. Since similes may perform various narrative functions (see Harding, 2017; Moon, 2011), it is worth considering their potential for emergent eventfulness in Cave’s writing. For example, as Nuttall (2018: 179) posits in her reading of Ballard’s The Drowned World, the explicitness of simile may serve to create an ‘awareness of what is real and what is imagined.’ Building on Browse’s (2014) application of Cognitive Grammar to metaphor analysis, Nuttall discusses novel similes that depict ‘the motion of an entity that is unfamiliar, fantastical or highly abstract’ (2018: 171) and argues that such similes, by invoking literary reference points outside the ground of the novel’s discourse, disrupt familiar apprehension of metaphorical connections, rendering the evoked conceptual domains themselves dynamic. Although Nuttall suggests that the disruptive, even surreal, character of Ballard’s similes is integrally linked with the genre conventions of science fiction, we would like to investigate whether similar patterns are present in our data, given that Cave’s writing often has an abstract and/or supernatural quality to it. Correspondingly, motion events are an obvious category of interest in our analysis, but we will also propose that descriptions of psychological processes may conjure up notably dynamic embodied effects as they unfold over time.
With a view to this, it is prudent to briefly address the nature of narrativity in lyric texts. As Hühn points out, lyric texts cannot be called narrative in the strict sense, but they do ‘refer to temporal sequences, changes of state and developments which comprise constitutive narrative elements, such as characters endowed with mental interiority, successions of happenings or actions motivated and made coherent by intentions, desires or anxieties and accompanied by emotions.’ (Hühn, 2016: 10) The notion of an event as an experiential ‘change of state’ seems particularly helpful to analysing the meaning-making potential of dynamic similes in our data. Namely fleeting and/or very compact narrative sequences require that the recipient activates their ‘narrative competence to fill in gaps and supply missing or merely implied connections by associating the appropriate conventional schemata or stereotypical procedural patterns’ (Hühn, 2016: 11). This resonates with Dancygier’s view of simile as creating discourse dynamicity through comparing the ‘experiential similarity’ of two situations (2022: 218). To what extent Cave’s similes can plausibly be said to invite this – keeping Morini’s (2013) critical remarks on the ‘attentive’ listener of song lyrics in mind – is an intriguing question to follow up in the analysis.
For a more granular analysis of the verb phrase in similative meaning and its potential embodied effects in relation to events, how a verb profiles a process (Langacker, 2008: 100) is of particular interest. In his Cognitive Grammar, Langacker compares the apprehension of an event to watching a film, as opposed to still images: ‘An event’s conception is continuous rather than discrete, even though each time-slice consists of a simplex relationship. These component relationships—referred to as
At the clausal level, Langacker (2008: 361) views the verb and its complements as together forming the ‘grounded structure’, which provides a situational context for processing the conceptual content (Langacker, 2008: 259). Alongside the focal participants, a verb’s meaning may involve ‘additional schematic entities that are sufficiently salient to function as elaboration sites.’ (Langacker, 2008: 360) An elaboration site may be inherent to the kind of relationship a verb profiles, for example complementing the verb put with the location to which an entity is moved. Further, in the case of modification with adverbial expressions, which can be paralleled with the like-part of simile, elaboration may occur in a broad range of ways, conveying ‘time, place, manner, degree, rate, means, reason, purpose, attitude, epistemic judgment’ (Langacker, 2008: 362). Due to this range of possibilities, it remains to be determined with each particular example whether the adverbial expression is a focal participant in the construction or not. The verb is decisive here in that its ‘internal structure is largely responsible for the grammatical organization of the clause as a whole. It determines not only how the situation is construed but also what remains to be expressed by other means.’ (Langacker, 2008: 400) Therefore, since our designation ‘dynamic simile’ initially only excludes stative verbs, it will be of interest to pay attention to the temporal aspect of the processes under scrutiny. To achieve this, we follow Croft’s (Croft, 2012: 33–35) view of aspectual construal through a focus on duration and boundedness;
3. Data and method
Nick Cave’s lyrics were collected from the official online lyric resource 2 and saved as plain text files. The text files were then uploaded to the CQPWeb interface (Hardie, 2012) and automatically part-of-speech (POS) tagged, creating a 55,661-word corpus of Cave’s song lyrics.
All examples of like used as a preposition were extracted from the corpus using POS-tagging.
3
The resulting concordance lines (N = 148) were then categorised into simile and non-simile following our previously defined criteria for simile; examples (1-2) illustrate cases of simile (1) and non-simile (2) found in the data. (1) ‘His blood spilled across the bar (2) ‘meanwhile Larry made up names for the ladies
A second round of categorisation was then performed to locate similes including a dynamic (non-stative) verb phrase (VP). Each simile construction was categorised according to the following criteria:4,5 (a) Does the simile contain a VP in the part preceding the like-phrase (3), in the like-phrase (4), both (5), or neither (6). (b) Is the VP dynamic (7) or not (8). (3) ‘The queen’s heart (4) ‘Little white clouds like (5) ‘I (6) ‘Hair of gold and lips like cherries.’ (Cave, 1986) (7) ‘While the bones of our child (8) ‘The moon was turned toward me like a platter made of gold’ (Cave, 1988a)
As the qualitative analysis of this study focuses on the contribution the dynamic VP makes to the realisation of embodied experiences in the similes of Cave, it was deemed fruitful to gain insights regarding the semantic content conveyed by the VPs. To achieve this, the dynamic VPs identified in the similes were semantically tagged using the automated online USAS semantic tagger (Wilson and Rayson, 1993).6,7
Of course, polysemy and contextual meaning entail that automated semantic tagging is not precise. Therefore, the results of the semantic tagging were manually checked by the authors and corrected in cases where the available contextual information provided information as to the relevant semantic category of the VP. An example of this difficulty is shown in (9): (9) ‘A jittery TV glowing white like fire’ (Cave, 2016b).
The automated semantic tagger categorised glowing as ‘O4.3’ (‘O. Substances, Materials, Objects and Equipment - Colour and colour patterns’); however, the authors judge that in the given context, the VP glowing conveys a visual experience rather than only denoting the properties of an object and therefore the decision was taken to categorise glowing in this case as ‘X3.4’ (‘X. Psychological Actions, States and Processes - Sensory: sight’).
In order to answer the first two research questions pertaining to Cave’s simile use when compared to other genres, two further datasets were created using sub-corpora from the British National Corpus XML edition: ‘Fiction & Verse’ and ‘Newspapers’. The two sub-corpora were selected as they provide both a stylistically similar and dissimilar comparison. 8 Moreover, as the BNC is a 112-million-word corpus it was necessary to create random samples of the data. 9 For each sub-corpus, a random sample of 150 utterances containing like used as a preposition was created. The same categorisations were then performed as previously described for the Cave dataset.
In the next section of the article, we provide an overview and discussion of the quantitative findings 10 , which is followed in sections 4.2 and 4.3 with a close qualitative analysis of how the VP in Cave’s similes contributes to a realisation of embodied experience.
4. Results and analysis
4.1 Quantitative results and analysis
‘Like’ simile frequency and use of dynamic VP in Cave, BNC Fiction & Verse, and BNC Newspapers.
It is, however, important to recall that the BNC data consists of random samples as the raw hits of like used as a preposition in the BNC are extremely numerous. Therefore, the findings presented should only be interpreted as an indication of simile frequency. Furthermore, we judge that the size of the datasets used in this analysis render statistical analysis unsuitable. However, despite the small size of the datasets, it is still of interest to compare the frequencies of simile use.
As described previously in the method section, the dynamic head verbs of similes were semantically categorised. The results of this procedure are shown in Figures 1 and 2.
11
As Figure 1 illustrates, the head verbs in Cave’s dynamic similes involve 12 unique semantic categories. In comparison, eight semantic categories were identified for the dynamic head verbs in the similes of the BNC Fiction & Verse sub-corpus, while for BNC Newspapers, only six unique semantic categories were identified. However, it should be caveated again that the results for the BNC sub-corpora are based on small random samples and, therefore, the results may not hold if the whole material were to be analysed. Distribution of the semantic categories of dynamic VPs in Cave, BNC Fiction & Verse and BNC Newspapers. Overview of semantic categories M & X in Cave, BNC Fiction & Verse and BNC Newspapers.
The data, however, do illustrate that two categories dominate for both Cave and the BNC Fiction & Verse, ‘M’ (‘movement, location, travel and transport’) and ‘X’ (‘psychological actions, states and processes’). This finding is of interest as these two datasets are those closest in terms of genre. The BNC Newspaper sub-corpus shows a somewhat different distribution with categories ‘A’ (‘general and abstract terms’) and ‘X’ proving the most frequently used semantic categories. Again, this finding makes sense when the genre of this sub-corpus is considered; news reporting regularly refers to quite abstract concepts and events, or, following Wikberg (2008), function to convey sensory experiences which are hard or even impossible to communicate using other means. The results of this initial semantic classification were used to select the categories ‘M’ and ‘X’ for closer analysis in the qualitative discussion (sections 4.2 and 4.3 respectively) concerning the function of dynamic VPs in Cave’s similes.
Figure 2 illustrates the more detailed semantic categories identified within the main categories ‘M’ and ‘X’ for all three investigated datasets. Category ‘M’ contains concepts with a relationship to movement and place, ranging from general movement (‘M1’) to specific methods of transportation by land, water and air (‘M3’ – ‘M5’, respectively). This group also contains movement that is caused from the application of force (‘M2’) and reference to location and direction (‘M6’). While there is a similar distribution of the individual semantic categories within the main groups of ‘M’ for both Cave and the BNC Fiction & Verse, there are some differences. An example is the notably more frequent use of verbs that relate to the application of force used by Cave (‘M2’). For texts in the BNC Newspaper data, category ‘M’ is restricted to general movement verbs.
The second category we choose to focus on (‘X’) contains concepts related to psychological and perceptual processes. In all our datasets, the most commonly used verbs categorised within ‘X’ belong to those expressing perceptual experiences (‘X3’) and its subcategories that represent the five senses: taste (‘X3.1’), sound (‘X3.2’), touch (‘X3.3’), sight (‘X3.4’) and smell (‘X3.5’). As Figure 2 shows, there is again a similar distribution for the Cave similes and those found in the BNC Fiction & Verse, with introspective (‘X3’) or perceptual verbs (‘X3.1 – X3.5’) being the most common. Interestingly, in all datasets the perceptual verbs are restricted to those expressing sound or visual experience – Cave does, however, use one verb related to smell.
In comparison to the BNC Newspaper dataset, both the Cave and BNC Fiction & Verse datasets use a wider range of perceptual verbs, particularly pertaining to audition, including the more iconic, or onomatopoetic, beating, churring and banging; this iconicity arguably conveys another layer of physical experience to the following simile constructions. The difference in physicality and figuration created by the use of sensory verbs related to audition is exemplified in (10–12). (10) ‘Archaeologists are sceptical about a claim that sounds like a Hollywood dream’ (BNC Newspapers AAT). (11) ‘Invisible grasshoppers were churring like little birds’ (BNC Fiction & Verse A0N). (12) ‘And your heart that is banging and beating and banging like a gong’ (Cave, 2004c).
Example (10) taken from the Newspaper sub-corpus illustrates the point made by Wikberg (2008): it would be very difficult indeed to convey the content linguistically without using the perceptual verb ‘sound’. In contrast, example (11), through the use of the more specific and onomatopoetic verb churr, conveys a more ‘perceptual experience’ than would have been achieved with the standard auditory verb, ‘sound’. Interestingly, although the simile used by Cave in (12) is constructed from a very conventional metaphor/simile blend, ‘a heart beating like a drum’, the alternation between bang and beat together with the somewhat novel gong instead of ‘drum’ creates an arguably more creative and iconic simile.
4.2 Qualitative analysis of similes involving movement
In the similes representing the most prominent movement category, ‘M1’ (‘coming and going’), the verbs come and walk appear several times. A selection of concordance lines reveals considerable variation in their usage: (13) ‘My aim is to hit this Miss And I’m movin’ in (I’m movin’ in) (14) ‘Maybe it was you or maybe it was me? You (15) ‘To make bright and clear your path And to (16) ‘I was
In (13) and (16), the verbs appear to be used in their basic sense, entering a location (13) and moving about in a location (16). When including the like-phrase as per the procedural approach (Walaszewska, 2013), a more fine-grained concept of the process that is profiled emerges. In both cases, the phrase indicates the manner of moving, although not in a plainly obvious way, which initially contrasts with Wikberg’s (2008: 139) observation about the like-phrase adding ‘further precision’ to the activity conveyed. In (13), the type of movement implied by the text-external referent ‘Lazarus’ is rather ambiguous, even when the source text for the reference is known. However, the directionality of ‘at her’ combined with the framing offered in ‘[m]y aim is to hit this Miss’ adds the purpose of seduction to the manner of moving, ‘Lazarus from above’ then possibly conveying a sense of renewed energy in this regard. In (16), two types of illness are associated with the manner of walking, imposing notions of physical unwellness (‘like a leper’) and emergent erraticness (‘Coming down with some kind of nervous hysteria’). The latter is particularly processual due to the unbounded nature of ‘[c]oming down with’ (cf. Croft, 2012: 34), which in turn arguably intensifies the narrative quality of the scene, in line with Hühn’s (2016) view of eventfulness in poetry often involving changes in mental state. The processual narrative quality is also reinforced in the progressive form of walk. It narrows the scope of the lexical verb to a limited duration (cf. Langacker, 2008: 361), which together with the complements depicts a particular, evolving state at a particular time. These examples can be seen as descriptive, as per the characterisation of simile’s functions by Israel et al. (2004: 130), but the sensory richness also relies on apprehending more abstract components (‘Lazarus’, ‘some kind of nervous hysteria’).
However, it is debatable whether these details will have notable readerly effects in engaging with the songs aurally (cf. Morini, 2013), and it is worth considering more transparent examples as well for a fuller picture. The similes in (14) and (15) are quite conventional and uncomplicated syntactically, thus aligning with the idea that simile typically prompts an instantaneous mapping (Moder, 2008). In (14), an emotional effect is suggested by combining come with the preposition on and the noun phrase ‘punch in the heart’. While the noun ‘punch’ is commonly used to describe emotions as physical reactions, together with on it arguably adds a limitedness of duration to come, making the verb prominently convey a state change. As for (15), ‘walk like Christ’ is an allusion to a well-known passage from the Bible, ‘He that saith he abideth in him ought himself also so to walk, even as he walked’ (1 John 2:6) 12 , and the addition of manner ‘in grace and love’ disambiguates the meaning further. However, the like-phrase diminishes the potential dynamicity of walk (which is already rather residual due to the infinitive): it activates the metaphorical idea of life as a journey, where choosing a particular path suggests life philosophy more than actual movement. This is then an example of a Metaphorizing Simile (Goatly, 2011), and the lines indeed reflect an overarching theme of providence as guided movement in the song (‘Into My Arms’).
Delving deeper into the data in category ‘M1’, a similarly accessible example involving the verb tumble is found in ‘Abattoir Blues’ (2004a): I kissed you once. I kissed you again My heart it Do you feel what I feel, dear?
Here, the simile clearly conveys an internal experience; the ‘I feel’ in the line following the simile makes this very apparent. Moreover, the beginning of the sentence containing the simile is quite conventional as regards the process profiled in the verb in conjunction with ‘heart’. The phrase ‘My heart […] tumbled’ does not arguably even need to be illuminated with a comparison. Considering that, the unexpectedness of the source ‘stock exchange’ seems meaningful. Coupled with the counting of kisses in the preceding line, the depiction of the internal state becomes markedly less ‘raw’ in an embodied sense and more aligned with reason. The intransitivity of the verb points to a thematic, as opposed to agentive, process (Langacker, 2008: 307), in which both ‘heart’ and ‘stock exchange’ can be conceptualised as movers. In Cognitive Grammar, theme ‘subsumes a number of “passive” semantic roles […]: patient, mover, experiencer, and zero.’ (Langacker, 2008: 366) The passive theme in the quoted lines contrasts with the implied calculative actions constitutive of the stock exchange, which showcases how dissimilarity may sometimes be instrumental to similative meaning and may even carry ideological significance through the nature of the contrast (cf. Moon, 2011). Indeed, the same contradiction can be seen in the title ‘Abattoir Blues’, which combines associations to slaughtering with a state of mind characterised by passivity. 13 The song as a whole contrasts acts of ‘[m]ass extinction’ with love as an escape from the harsh reality, and though it is not specified whether the protagonist has actively contributed to the former, the juxtaposition connects with Eaglestone’s (2009) observation about immoral decisions and softer themes often being intermingled in Cave’s lyrics.
By contrast, in ‘Rings of Saturn’ (2016a) the verb dangle is used to describe an intimate scene in a way that requires quite some effort to decipher: And now she’s jumping up with her leaping brain, stepping over heaps of sleeping children, […] up and out of the bed and down the hall where she stops for a moment and turns and says ‘Are you still here?’ and then reaches high and
The verb belongs to category ‘M2’, including actions that involve the application of force (‘putting, taking, pulling, pushing’). However, despite the verse depicting movement away from the bed that the speaker and their lover have previously occupied (‘she’s jumping up’, ‘out of the bed and down the hall’), it is unclear whether ‘dangles herself’ conveys any actual movement, given the cosmic expansion of the scene in the like-phrase. While the image of ‘the ring of Saturn’ as something tangible is captivating and as such testifies to the Grounds’ capacity to manipulate the recipient’s attention (Tartakovsky et al., 2019: 187), the example also relates back to Goatly’s (2011: 195) point about interpretive challenges emerging if the Grounds appear inconclusive – it is not obvious how reaching upward (into space) resembles ‘a child’s dream’. Here, Cave’s own gloss in The Red Hand Files offers an illuminating perspective. Cave characterises the lyrics on the album as ‘ambiguous’ and ‘unguarded’ and states that in this verse the woman ‘ascends into the melancholy splendour of the cosmos […] in recognition of both the joy and sorrow of womanhood.’ (2019a: n.p.) It would seem, then, that the act of getting up post-sex coincides with a complex emotional elation, captured in ‘a child’s dream’ turning the phrase inward and towards earthly responsibilities and ‘the ring of Saturn’ upward and away from mundaneness. Although quite multilayered, the simile’s expenditure of force via dangle is arguably made more accessible through the music: Cave explains that he enjoys performing the song live ‘because the rising, building chords take us to a place of collective bliss’, thus mirroring the song’s physically expressed theme of ‘transcendent love’ (2019a: n.p.).
4.3 Qualitative analysis of similes involving psychological actions and processes
As noted by Barfield (2013), Eaglestone (2009) and others, Cave’s songs often feature dark themes and fictional characters with questionable motives. This is a tendency that is seen in several of the similes in the category ‘X’, pertaining to psychological actions and processes. A notable example comes from one of Cave’s most well-known songs, ‘Red Right Hand’ (1994): Take a little walk to the edge of town Go across the tracks Where the viaduct As it shifts and cracks
Here, physical movement (‘walk’, ‘go across’) is combined with the ‘visual’ verb loom to compare the viaduct with a living creature: its shape associates visually with a bird in flight. The dictionary meanings of loom give rise to both a menacing appearance and something less tangible that causes concern, such as a future event. The use of sound patterning supports both the visual and the more global meaning: ‘looms’ rhyming with ‘doom’ in the like-phrase transfers the source’s menacing agentive force to the inanimate target. This is continued in the noises the viaduct makes, pointing to a potential future event (‘it shifts and cracks’). The foreboding atmosphere is further accentuated by the fact that to actually compare the viaduct with a bird entails that the bird is in flight or at least has spread its wings – not only is there a suggestion of movement, but ‘bird of doom’ activates the idea of a bird circling over its prey. This could be linked to Nuttall’s discussion of novel similes depicting motion by an abstract or unfamiliar entity through text-external reference (2018: 171). In this case, the implied movement of a predator results, by way of schema activation building on experiential similarity (Dancygier, 2022), in an atmospheric animation of the urban setting, ultimately suggesting that an unspecified threat is active.
While the use of the verb loom in ‘Red Right Hand’ suggests that danger is lingering, in the song ‘Worm Tamer’ (2010), written for Cave’s side project Grinderman, a verb phrase conveying a very short duration indicates threat: ‘You know they call my girl The Worm Tamer / She
Continuing the somewhat ambiguous use of verbs in conveying psychological processes, ‘Slowly Goes The Night’ (1988b) paints a scene from a relationship using the verb trace: So slowly goes the night I
Like the map of some forbidden
land
In this example, the simile initially seems to convey a physical action. However, before the occurrence of the simile, the words ‘lover’ and ‘goodbye’ have established the theme of lost love, while ‘night’ in the preceding line allows the recipient to interpret the target of the simile, ‘the print of your body’, as the print left on the bed by the absent lover’s body. Even if the action is clear with regard to the target, it is not obvious in what way tracing ‘the map of some forbidden land’ is comparable to it. Considering the immediate scope (Langacker, 2008: 63) of ‘the print of your body’ and ‘the map of some forbidden land’ in relation to the implied maximal scope can help illuminate the figurative comparison. Namely the of-structure in both points to the fact that the source and the target are representations of things that are not present within the scenario, that is the lover’s body and the unattainable land portrayed by the map, and thus a mental focus becomes prominent. In other words, these gestures towards the maximal scope draw attention to how the utterance suggests a state of absence, allowing for a subtle tensional balance (Fishelov, 1993) in the simile to be attained. Crucially, the verb trace highlights the temporality of experiencing that absence.
In ‘Wide Lovely Eyes’ (2013), a very similar physical experience of time is depicted: You wave and say goodbye The night expands I am expanding I
While the simile simply conveys a visual observation, that of the lover’s hands during and/or after waving, touch is also residually suggested through the like-phrase: the delicate nature of a butterfly landing activates the idea of a fluttery motion and a very light touch, as opposed to a quick, single-motion change. This is also suggested by the progressive form of the verb land and the fact that the word ‘wave’ has been repeated (in its different senses) several times in the song: the moment seems to last, as if in slow motion. Furthermore, the effect is arguably primed by the phrase ‘I am expanding’ in the previous line 14 : as abstract as this process appears on its own, it quite clearly carries an emotional meaning due to being presented as a response to the lover’s wave. There is also an interesting contrast between the protagonist and their surroundings in this line: ‘The night expands’ being in present simple form, rather than progressive form, brings into focus how the inner experiences (‘I am expanding’, ‘like butterflies landing’) are particularly processual (durative and unbounded, Croft, 2012: 35), and as such more intensely dynamic. This also makes for an interesting contrast to the use of trace in ‘Slowly Goes The Night’, and indeed ‘Wide Lovely Eyes’ is not about loss of love but about sharing the fleetingness of moments with a loved one.
5. Concluding remarks
Our investigation demonstrated that in comparison to other genres of both similar and dissimilar language, like-simile is used by Cave at notably higher rates. We could also show through semi-automated semantic analysis that the verb phrases used by Cave tend to belong to domains of (physical) movement and psychological and perceptual processes. This finding supports previous scholarship that has noted the prominence of embodied experience in Cave’s song lyrics. Cave’s similes were also found to display considerable range in the use of dynamic head verbs compared to the two sub-corpora of the BNC. Although not investigated by the present study, further research with a diachronic perspective concerning Cave’s similes, or other language features, could be of interest as Cave himself has discussed a changing lyrical style during his career thus far.
By focusing on the verb phrase in like-simile, the analysis showcased how dynamicity may be an integral part of similative meaning. Noteworthy readerly effects may be gleaned not only from the choice of verb but also its grammatical form and the ‘additional schematic entities’ (Langacker, 2008) it is accompanied by in the clausal structure, for example details regarding duration, direction and subjective perspective that flesh out the narrative scenario. While the like-phrase often functioned as an adverbial specifying manner in Cave’s similes, in line with Wikberg’s (2008) findings, the combination of verb and the expression of manner frequently required interpretive efforts aided by the context of the song. The discussion also made evident how similes containing a dynamic verb phrase may produce notably different effects depending on the situational framing. In a handful of Cave’s similes, the use of motion verbs pointed to relatively clear narrative events (such as physically approaching a person), but in most cases the verb phrases contributed to presenting various internal processes in an embodied way. Though musical aspects were not included in the analysis due to the article’s limited scope, on several occasions we observed dynamic similes’ potential for resonating with the recipient at an experiential level. We would suggest that exploring the iconic relationship between such effects and the dynamics of music is a potentially rich future avenue for stylistic research on song lyrics.
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.