1. Introduction

1Winged words, famous quotations or dicta differ considerably in their etymology and modality. While proverbs often cannot be traced back to their origins and are primarily orally transmitted, winged words can be traced back to a source and are part of the written tradition. Thus, at least the process of coinage of winged words is not as difficult to fathom as it is in the case of proverbs. In this line of thought it can be assumed that winged words had propagated primarily via their original source and through reading and reciting. Due to the Judeo-Christian and Greek-Latin cultural heritage, a plethora of phrasemes in the languages of the Western world derive from stories and narratives that have evolved with European civilisation and are still culturally highly consistent, even if they seem outdated because they are centuries old, like the Bible or literary classics. In this contribution winged word is used as phraseological umbrella term for all other types of famous quotations / dicta / catchphrases / slogans / sententiae / maxims / aphorisms / allusions / fragments of quotations etc., i.e. phrasemes that are retraceable to a source or an author, as for example Burger [2015] does. Although there are some terminological and typological issues, as addressed by Schütz [2023: 114-115; 117], the term appears to adequately convey the idea of a famous quotation that becomes more popular outside of its domain of origin.

3Based on Heylighen & Chielens [2009], Villers [2022] formulates a series of fundamental propagation boosters for proverbs, e.g. ornatus, novelty, usefulness, simplicity, veracity, conformity, and prestige. Consequently, by measuring the characteristics of phrasemes we can establish a hierarchy of usage. In memetic terms, a successful phraseme must spread to as many hosts as possible but is itself aligned with the extralinguistic world, i.e. it must be coherent with the conceptual and cultural value hierarchies in which its coinage is embedded.

4So far, the role of values that winged words correspond to has not been addressed, which is why this paper aims to fill this gap by reflecting on the underlying values and the implications for phraseme coinage and propagation. The first section of the article combines a pragmalinguistic view on phraseology with a review of cross-disciplinary evidence favouring the interdependence of perception, cognition and language regarding the following topics: (1) bottom-up and top-down processing in visual and auditory perception (Gaspelin & Luck [2018]; Angosto et al. [2013]; Riener [2019]), (2) brain activity in word processing and meaning processing (Pulvermüller [2002, 2005]; Hauk [2004]; Rizzolatti & Craighero [2004]), (3) embodiment and learning (Goodwin [2003]; Gee [1995]).

6In the second section, a corpus study is carried out for four iconic winged words and (potential) modifications or patterns focusing on their intertextuality and the relationship between the figurative elements and the corresponding value and meaning structures: To be or not to be, that is the question by William Shakespeare, Das also war des Pudels Kern! By Johann Wolfgang von Goethe, Nel mezzo del cammin di nostra vita by Dante Alighieri, and one Bibleism, an eye for an eye. One question arises: how stable and prevalent is the canonical form and to what degree does it tend to be modified?

2. Values and meanings: how do people fundamentally orient themselves perceptually and linguistically?

7This chapter undertakes an interdisciplinary excursus into the peculiarities of human cognition and its interweavement with language and perception. This is an approach to two subjects of two usually separate academic disciplines; however, as it is supposed, they intersect in one fundamental aspect. The first one issues the particular pragmatic potential of phrasemes, or as Schafroth [2013: 187] puts it, we typically want to achieve something when we speak, i.e. when we act linguistically, and phrasemes appear to facilitate that, perhaps because we do not directly articulate what we mean assuming the listener will understand. The second one regards the rather general question of which processes govern how humans orient themselves in the world from the perspective of cognitive psychology. The common core is the question of what we are oriented towards when we act, including linguistically.

9Imagine a toddler crawling from one corner of a big room to the opposite one because a parent calls for him. He knows his environment, the spatial dimensions, the objects in it. On his way to his destination, he encounters several difficulties: plants, a sofa, a table, toys etc. are blocking his passage. He does not exactly perceive the chair in his way as an object to sit down, nor does he bother playing with the toys nor do anything else but continue to crawl in the direction of his parent. Instead, he notices obstacles to navigate around in order to reach his goal.

10Cognitive psychology examines “the internal processes involved in making sense of the environment and deciding on appropriate action” and includes attention, perception, learning, memory, language, problem solving, reasoning and thinking (Eysenck & Keane [2020: 1]). Values and meaning seem to form a joint connection between the individual, its environment and other people. To analyse phrasemes holistically we must try to develop a deeper understanding of perceptual and cognitive processes involved in communication, and cognitive psychology shares this holistic endeavour.

11“Perception is the study of how sensory information is processed into perceptual experience” (Twedt & Proffitt [2018: 1]), so to perform actions people need to filter and process sensory input from the external environment. Within the scope of this article, visual and auditorial perception are singled out because of their crucial importance to language processing, since visual perception permits humans to read texts as auditorial perception does to comprehend speech.

12The traditional notion of visual perception is the “bottom-up processing” of light reflecting on surfaces as sensory input through the eye and then processed by the brain (cf. Riener [2019: 267]). Proffitt [2006: 110] specifies that visual perception can be influenced by people’s purposes, their physiological state and emotions and thus is not exclusively visual:

Perceptions are embodied; they relate body and goals to the opportunities and costs of acting in the environment. Here are some examples: Under constant viewing conditions, the apparent incline of hills increases when people are tired or encumbered by wearing a heavy backpack; hills also appear steeper to people who are in poor physical condition or who are elderly and in declining health, compared with those who are young, healthy, and fit. Similarly, apparent distance increases when the observer is encumbered by a backpack or throwing a heavy ball. When one is standing on a high balcony, the apparent distance to the ground is positively correlated with one’s fear of heights.

13Perception thereby not only makes it possible to perform actions, but it is relating possible actions to energy consumption. Riener [2019: 269-270] compiles further studies that support Proffitt’s “embodied approach” to perception, including steeper perception of hill slants by impaired people, body width influencing the estimation of the size of apertures, holding out an arm influencing estimated width of apertures and people rescaling spatial information to the body. He sums up that “perceptual representations are constructed to be applied to future action and thus are grounded in the body that will be performing the action” (Riener [2019: 270]). Spatial orientation can impact cognition. There are culture specific conceptual differences that manifest linguistically. Regarding the conceptualisation of time and the duration of an event, English uses metaphors of linear distance (e.g., a long time, like a long rope), however, a Greek speaker is likely to use amount metaphors (e.g., a lot of time, like a lot of water) (Casasanto & Bottini [2014: 141]).

14What Riener [2019] and Proffitt [2006] describe, are examples of “top-down processing”, the phenomenon that people’s past expectations, knowledge and experiences can influence their perception. But “top-down” does not imply awareness or intentionality as Gaspelin & Luck [2018: 2] argue with examples of visual processing and auditory processing. Consider the figure below:

15You certainly encounter no difficulties in reading the sentence. Gaspelin & Luck [2018: 2] explain:

The “A” in “CAT” and the “H” in “THE” are exactly the same shape, but top-down knowledge of common English words allows experienced readers to read the words as “CAT” and “THE” without even realizing that they could be seen as “CHT” and “TAE”.

16Another example of visual top-down processing in reading is presented in Figure 2:

17If you read it as “Paris in the spring”, try reading again. The traditional hypothesis was that processing occurs serially, when in fact several processes can occur simultaneously (parallel processing) or they can overlap (cascade processing) (cf. Eysenck & Keane [2020: 4]). An example of top-down auditory perception is the phonemic restoration effect: an auditory illusion in which listeners restore missing parts of speech signals where a meaningless noise occurs (Gaspelin & Luck [2018: 2]; Samuel [1981: 474]). There are a few interesting similarities between reading and auditory perception: adults read “relatively fast (on the order of 250-350 wpm (words per minute) for most skilled readers), and listening can approach but probably not reach this rate” (Rayner & Clifton [2009: 6]) and “language processing is fast and largely incremental, in the sense that substantial syntactic, semantic, and pragmatic processing of a word is done while the eyes are fixated on that word or while that word is being heard” (Rayner & Clifton [2009: 4]). Furthermore, before children learn to read, they learn to understand speech, which means that there are processes involved in understanding that are also involved in reading; Farmer & Klein [1995: 487] found sufficient evidence that a temporal processing deficit may be the cause of dyslexia in a substantial number of cases. Temporal processing, or sequential processing, refers to “any processing procedure involving two or more stimuli presented non-simultaneously” (Farmer & Klein [1995: 462]).

19Evidence for a dynamic view of word meaning in the brain was found before by Hauk et al. [2004: 301] who spotted specific patterns in brain activity: “action words from different semantic subcategories (referring to movement of parts of the face, arm, or leg) activate the motor cortex in a somatotopic fashion that overlaps in premotor and motor cortex with the activation pattern observed for actual movements of the relevant body parts.” Pulvermüller [2002: 581] affirms the reciprocal connection between the brain system for action and language: “Action meaning seems to be not only necessary, but also highly relevant for language.”

20Rizzolatti & Craighero [2004: 172] portray how fundamental motor action is for people finding that when “an individual sees an action done by another individual, neurons that represent that action are activated in the observer’s premotor cortex”, and the “motor representation of the observed action corresponds to that which is spontaneously generated during active action and whose outcome is known to the acting individual.” The so-called mirror neurons are activated in both cases, performing actions and observing them. Mirror neurons are believed to have an impact in understanding actions of others as well as imitating them (cf. Eysenck & Keane [2020: 161-163]).

21Goodwin [2003] analyses videotapes of archaeologists learning how to excavate from an embodiment perspective to draw conclusions about the interplay of body, language, cognition and social organisation. He states that “[t]he positioning, actions, and orientation of the body in the environment are crucial to how participants understand what is happening and build action together” [2003: 20]. In this situation cognition is highly socially coordinated. Atkinson [2010: 603] claims that “cognition is shared and distributed” and sees co-cognition supported by the “human ‘mirror neuron system’.” Goodwin’s notion corresponds in part to Gee’s [1995: 337] public principle:

The meanings of the parts of new systems, whether words, visual symbols, actions, or objects, must initially be rendered public and overt, so that the learner can see the connections between the signs and their interpretations. And this is done, in first language acquisition and other forms of learning, by the ways in which words, actions, and social interaction are integrally intertwined.

22By studying first language acquisition, he formulates these principles to develop a first language acquisition-based theory of learning but claims their general applicability to other sorts of learning [1995: 331]. The idea is indeed quite consistent with other theories; for example, it is precisely what Grice [1975] describes as cooperative effort of speaker and listener to understand each other and what psychologists call “common ground” (Eysenck & Keane [2020: 484]), which is shared information and knowledge about the world.

23Our perception helps us to orient ourselves in the physical environment, it is not possible to process the entirety of sensory information there is in a given situation: instead, we learn how to align ourselves with reality, rather than filtering everything there is, we are guided by meaning in the context of performing an action. Analogous to the human orientation for action in the spatial dimension, the pragmatic view can be extended to language, because language processing shows that people are oriented towards meaning and action; performing abstract linguistic actions helps us to navigate and solve more complex situations and problems in the world.

3. Intertextuality as gateway to the cultural nexus

24Pham [2014: 2] states that the term intertextuality was created in the context of post structural literary theory, in terms of mimesis, imitatio, quotations and parodies, investigations on intertextual connections of texts are retraceable to classical antiquity. She analyses intertextual references to the Hamletian soliloquy to be or not to be, that is the question and, based on frame semantics, develops a model of intertextual frames, cognitive knowledge representations in which all the knowledge necessary for the production and / or reception of intertextual references is stored [2014: 11]. Arguing for an operational definition of intertextuality as a text-descriptive approach that takes intertextual references as a criterion of difference between texts, Pham [2014: 46] views intertextuality as gradable, since texts can be categorised as texts whose production and reception do not depend on other texts, and as intertextual texts whose production and reception depend to varying degrees on one or more other texts. That gets support from the empirical findings in chapter 4 that indicate gradual differences in how winged words and modifications are reproduced.

25There are at least four connecting factors to be considered for intertextuality in the context of phraseological studies:

• The source

• Domain specificity

• Intertextual motivation

• Communicative functions

26The first characteristic is purely technical and regards the identification of the source. A winged word is a part of speech that is reproduced outside of the original source, and thus establishes a connection to the source text it originated in. Additionally, it is of interest what structural portion of the original is reproduced. The second aspect refers to the fact that winged words, and phrasemes in general, can be very specific regarding their domain of origin, but also culturally and socially, which plays into the degree to which they are institutionalised. Intertextual motivation is an adaptation from Dobrovol’skij & Piirainen’s [2009: 17-19] study on idioms in which intertextual motivation, along with metaphoric and symbolic motivation, is classified as one of three types of synchronic semantic motivation. The fourth factor regards the pragmatic potential of winged words, and particularly the communicative functions regarding rhetoric / stylistic elements, formal and prosodic aspects, production cost / reception effort reduction, self-expression and speaker-listener organisation. Each of the aspects (1-4) is explained in a more detailed fashion in the brief following subchapters.

3.1. The source

3.2. Domain specificity and acceptance

29After the coinage, a phraseme enters the lexicalisation phase: “The next stage in the history of a lexeme is when the nonce formation starts to be accepted by other speakers as a known lexical item” (Bauer [1983: 48]). The acceptance of winged words at least partly stems from the fact that a community of people values some texts over others and hence a hierarchy of texts and corresponding values differing in their importance for the collective is established. This is reflected for example by the number of books sold, the diffusion in different languages or in the fact that some authors or works are studied rather extensively in education institutions, while others are not. In the case of winged words, the devil lies in the details since winged words are often domain specific and their diffusion can thus be limited. Moon [1998: 7] acknowledges in general that “FEIs [fixed expressions and idioms] may be localized within certain sections of a language community, and peculiar to certain varieties or domains.”

30The beginning of intellectual interest in quotations and their codification in collections and dictionaries can be dated back to France, when historian and bibliographer Édouard Fournier published L’Esprit des autres (1855) and L’Esprit dans l’histoire, recherches et curiosités sur les mots historiques (1857). The coinage of the term “winged word” followed in 1864, very much a winged word itself – referring to the Homerian epiteth (gr.) ἔπεα πτερόεντα (epea pteroenta), but probably mistranslated (cf. Schütz [2023: 114]) – and at the same time the title of a collection of quotations by German philologist Georg Büchmann (Geflügelte Worte, Der Zitatenschatz des Deutschen Volkes). In 1895 Giuseppe Fumagalli published his collection of quotations Chi l’ha detto?, followed by Thomas Benfield Harbottle’s Dictionary of Quotations (Classical) in 1897 and by Philip Hugh Dalbiac’s Dictionary of Quotations (English) in 1908. These publications prove that at least amongst intellectuals winged words gained some degree of permeability, long before the digital revolution. At the same time, for certain winged words, such as those from literature, philosophy or science, one must presume a rather limited diffusion, simply because education was not accessible for most parts of the population until the twentieth century, when other texts, for example the bible, were substantially more accessible.

3.3. Intertextual motivation

32Due to phraseologisation, winged words can lose their intertextual association, whereas in some cases they maintain their intertextual motivation. The latter has been addressed by Dobrovol’skij & Piirainen from a cultural semiotic [1996] and a cognitive perspective [2009], though the authors focus on the intertextual motivation of idioms.

33“Any attempt to analyse figurative language by itself, without the inclusion of extralinguistic knowledge, is doomed to failure” (Dobrovol’skij & Piirainen [2022: V]). What Dobrovol’skij & Piirainen formulate here, can easily be extended to winged words, even if they do not contain figurative language: it would simply be doomed to failure not to include extralinguistic information in the analysis of winged words, because the fact that they can be intertextually motivated and often refer to shared extralinguistic knowledge is one of the main reasons they are used in the first place.

34Phrasemes that rely on imagery are potentially highly constrained by cultural and conceptual determinants: to cast pearls before swine (Matthew 7,6) can only be propagated if the culture-specific knowledge, the symbolic meaning of pearls as valuable is shared. Burger [2015: 13] calls biblical phrasemes such as to cast pearls before swine phraseological loan translations. Dobrovol’skij & Piirainen [2009: 35-36] identify lexical irregularities caused by archaic translations using swine instead of pig, or the existence of two variants in French, jeter des perles aux pourceaux and jeter des perles aux cochons (cf. ger. Perlen vor die Säue werfen uses Säue ‘female pigs’ and it. gettare perle ai porci uses porci ‘pigs’). Nonetheless, by translation the biblical texts gained a significant propagation boost. According to Dobrovol’skij & Piirainen’s [2009] model, phrasemes like to cast pearls before swine are cases of blending of intertextual and symbolic motivation. It contains the semantically autonomous component pearls, which is a symbol that appears outside the phrase with the same meaning (Dobrovol’skij & Piirainen [1996: 101]). The authors specify that it has no influence on the cognitive processing of an idiom if a speaker does not know the etymology of an idiom (Dobrovol’skij & Piirainen [1996: 104]), and if the speakers do not know the source text, the idiom is only metaphorically or symbolically motivated for them; if they are familiar with the text, the idiom is also intertextually motivated for them (Dobrovol’skij & Piirainen [2009: 32]). At this point some terminological caution is due since “[t]he motivation type based on textual knowledge must not be confused with “etymological knowledge” about the origin of an idiom from a certain well-known text” (Dobrovol’skij & Piirainen [2022: 115]), because it refers to “the knowledge of a prefabricated text that underlies the interpretation” [2022: 114-115].

3.4. Communicative functions

35Expressivity is often mentioned as one main characteristic of winged words, which refers on the one hand to the person who reproduces it and on the other hand to the reaction of the recipient. Reproducing a winged word can be used to demonstrate one’s own attitude toward the meaning, while on the other side organising the relationship with the recipient (Lüger [1999: 205]), which corresponds to the psychological notion of common ground between people. In other words, expressivity can cover a wide range of additional information, e.g. about emotional states, rhetorical skills or social affiliation. What Lanier [2002: 3] states through the example of Shakespeare can generally be assumed to be true for other language communities and the significance of particular authors: “Shakespeare symbolizes high art in general, the distinction between ‘Shakespeare’ and ‘popular culture’ epitomizes one of the great divides in the culture of the last century, the division between highbrow and lowbrow.” Burger [2015: 142] comments that winged words were already used in the nineteenth century to show social affiliation to the educated classes.

36However, communication is not a one-way street: we must appeal to common ground, we need shared knowledge for communicating successfully, especially in situations with time limitations. This is very consistent with the psychological view: people tend to know how to manoeuvre between an egocentric heuristic, roughly meaning overestimating what others know by judging off one’s own knowledge, and the common ground (Eysenck & Keane [2020: 484]). One way to overtly stress a winged word is by marking it meta-communicatively via mentioning the author in the direct context (Lüger [1999: 119]), supposedly to support an argument authoritatively. The main difference between written and spoken language in this regard is the lack of prosodic cues, which, can create ironic or humorous effects, when used competently, whereas in written language people can use formal cues like quotation marks.

37Studies on length of the most common proverbs show that they normally do not tend to be syntactically complex. While for English proverbs an average of 7 words is calculated by Mieder [2004: 7] and an average of 6 words by Villers [2014: 223], Lüger [1999: 103] notes that 52% of the cases analysed in a study of German proverbs by Grzybek [1995] are between 5 and 8 words long. He [1999: 103-104] asserts that sentence length correlates negatively with familiarity, as familiarity increases with decreasing sentence length and hypothesises this is due to higher plausibility and conciseness, facilitating memorisation and handing down, as well as the mnemonic effect of faster retrievability. It seems fair to extrapolate this data to winged words. The similarity between perceptual energy consumption management and language production is apparent: the notion of an economic use of language is made clear in the use of prefabricated linguistic expressions. As Stein [2007: 226] comments on prefabricated elements, they are ready to be used without any noteworthy cost of planning resources.

4. Winged words and patterns / modifications

38Clearly, phraseological modification is no black swan event as empiric evidence from different academic traditions has proven. By analysing modifications in relation to their canonical form, the variability of phrasemes can be quantified. According to the frequency principle, a high token frequency, a high number of occurrences in a corpus, correlates to the degree of cognitive entrenchment of a linguistic unit in a language community (cf. Ziem & Lasch [2013: 103]). From an onomasiological perspective it must be investigated on the canonical meaning of the winged words to then detect differences, e.g. semantic changes, in modern use and compared to modifications and patterns.

39Therefore, the main questions regarding the modification process are: how fixed are winged words? What possible slots can be identified and what is the lexical material filled in? In terms of the objective of the article, what can be said about an involvement of values in the propagation?

40The pattern to X or not to X is often mentioned in the context of snowclones, “some-assembly-required adaptable cliché frames for lazy journalists” (Pullum [2004]). Mellado-Blanco [2022: 10] comments on snowclones from a construction grammar perspective: “Although the term “snowclone” emerged outside of CxGs, this type of pattern can be identified as a semi-schematic construction, and thus it has not gone unnoticed by certain construction grammarians.” What Pullum calls “lazy” can certainly be viewed positively as effort efficient: reproducing a widely known pattern minimises production effort, but also processing effort, as discussed in the previous chapters. Schütz [2023: 117] reviews studies on the Italian literary winged word ed è subito sera and its modification in media and advertisement concluding: “Certainly, journalists and marketers intentionally create catchy headlines and slogans, hence premade structures minimise the productional effort and potentially maximise the effect.” However, simply put, people do not only use prefabricated material to “show off” or create intended effects, because simultaneously they appeal to the linguistic knowledge of others and communicate in good faith that the extralinguistic knowledge is shared.

41Based on the dynamics of perception, processing modifications of phrasemes theoretically leads to a collision between top-down and bottom-up processes in language perception: if we know the canonical form, syntactic and semantic priming can lead to inferring the whole phraseme, then a new element appears and we recognise that it is slightly different; we then recognise the new instantiation of a pattern based on a well-known phraseme. Philip [2008: 104] puts this process in phraseological terms in the context of modification by substitution to create “puns” arguing against applying Sinclair’s open-choice principle to phraseological chunks:

[…] open choice implies far more freedom than is actually available in this sort of phraseological manipulation, because whatever element is substituted, its meaning is always read in relation to the canonical phrase. The new element forces the reader to analyse the phrase both compositionally and non-compositionally, and the overall meaning is a combination of the old phrase and the new, and not a new phrase in its own right.

4.1. To be or not to be, that is the question

44The dichotomy to be or not to be refers to one philosophical question about human existence, the truth of which is then to be examined: How should people act with the knowledge of life’s suffering? Structurally, the winged word represents a special case of either / or disjunction, a special reduplication structure, as the first component is repeated but negated. The dichotomy between life and suicide is discussed argumentatively in Hamlet’s soliloquy. The question to be (A), or not to be (B) is contextually based on the meaning ‘Is escape in the form of suicide (B) a valid alternative to the evils of life (A)? The hopelessness obviously arises when a decision makes the alternative impossible and a life free of suffering seems inconceivable without a redemptive suicide. In terms of information structure, the quotation is divided into two parts, as the rheme (to be or not to be) is prefixed for emphasis. In the second part that is the question, the rhema is then taken up again as the thema (that) and emphasised once more, as it is made clear that it is a question or “the question”. In this respect, the second part does not contain new information and is used for emphasis. One could argue, the dichotomy serves to describe the object under consideration and is therefore to be seen as a paraphrase of the ensuing intellectual debate about two different choices, both of which are potentially associated with negative consequences. Yet, the interpretation of the quotation as a dualism of life and death in the sense of a dichotomy of good and evil represents a reduction of complexity, because the hopelessness rhetorically conveyed via the aporia is at the forefront of the protagonist’s monological reflection. Shore [2015: 129-131] proves that the verse is in fact a case of dual intertextuality analysing occurrences and use before 1600, and that it was not “loaded” as it is in the Shakespearean use, it was instead used in situations of indifference.

45To be or not to be is a prototypical disjunctive phraseme and it is used for communicating a value decision of choosing one thing over another (Shakespeare sets up the ultimate value decision there is: the value of life as opposed to death, thus to be or not to be).

Table 2. Top 20 types and token frequency of the slot filler in to X or not to X (left) and top 20 types and token frequency of the slot filler in to X or not to X that is the question (right)

Types X	Token frequency	Percentage	Types x	Token frequency	Percentage
be	7366	19.06%	be	490	16.68%
do	1616	4.18%	buy	50	1.72%
buy	837	2.17%	go	30	1.03%
use	502	1,30%	pay	23	0.79%
eat	358	0.93%	merge	22	0.76%
go	350	0.91%	do	21	0.72%
believe	345	0.89%	eat	20	0.69%
pay	313	0.81%	use	18	0.62%
act	271	0.70%	cut	17	0.58%
read	266	0.69%	sell	17	0.58%
see	260	0.67%	blog	15	0.52%
have	256	0.66%	believe	15	0.52%
share	223	0.58%	drink	15	0.52%
treat	214	0.55%	kill	15	0.52%
tell	204	0.53%	tweet	15	0.52%
vote	201	0.52%	bee	13	0.45%
give	187	0.48%	mask	13	0.45%
kill	181	0.47%	test	13	0.45%
join	181	0.47%	share	12	0.41%
cut	180	0.47%	vaccinate	12	0.41%

46The meaning of the pattern to X or not to X can be paraphrased as ‘How can X be achieved?’ or ‘What speaks in favour of X?’. The topos of the hopelessness of a situation expressed in the monologue by the aporia can usually no longer be identified after the transformation process. The disjunctive structure of the template does sometimes evoke reflection on the possible consequences of either choice; concluding from the results, particularly from slot fillers like everyday verbs to do, to buy, to use, to eat, to go, to pay etc., instead of expressing indecision it is usually used as introduction to argue in favour of one alternative. Consider this random example for to X or not to X:

(a) […] To Tilt Or Not To Tilt Your Portfolio Design. […] (enTenTen21, 38901157488)

47It is the title of an online article of Forbes Magazine, which discusses the pros and cons of an investment strategy called tilt / tilting, explaining that a “[…] “tilt” is industry slang for an investment strategy that overweighs a particular investment style” (Ferri [2014]). The following example is an instance of to X or not to X that is the question which is quite interesting:

(b) […] and I would not necessarily blame tanning salons themselves for causing skin cancer or skin problems. Even on cloudy days, you are at risk, so many dermatologists suggest that you always apply sunscreen whenever you go outside, even during the winter months. To tan or not to tan ... that is the question. It all depends on personal opinion, folks! […] (enTenTen21, 24713446095)

48The context indicates that the topic at hand is skin / health care in relation to tanning, tanning salons and the use of sunscreen. It is suggested that a lack thereof can lead to potential health risks such as skin cancer. In this sense, the example turns out to indirectly fit the original dichotomy of life and death, evoking the potential death from cancer through context, while the question of whether to tan itself seems to be a rather arbitrary everyday phenomenon.

4.2. Das also war des Pudels Kern!

49The winged word Das also war des Pudels Kern! (literally *that so was the poodle’s core) originating in Goethe’s Faust I is probably one of the author’s most commonly known phrasemes in the Germanophone world, used with the meaning ‘so that’s going on’ or ‘that is the real, previously hidden cause’. However, there are several lexicogrammatically reduced variants like etwas ist des Pudels Kern (something is the poodle’s core) or das ist des Pudels Kern (that is the poodle’s core). The syntactic feature of the preceding genitive attribute des remains, which is called a phraseological irregularity according to Stumpf [2015]. It seems plausible that this rare and striking structural feature, in addition to the idiomatic nature of the phrase and Goethe’s cultural and historical significance, has contributed to its stability.

50Stumpf [2015: 36] sees the form etwas ist des Pudels Kern as an example of an absolutely fixed phrase. He [2015: 155] empirically proves that phrasemes with preceding genitive attributes do not tend to allow positional variation by noting that in over half of the phrasemes analysed, the genitive attribute precedes in 95% of the cases.

51The phraseme is another example of dual intertextuality, by the special religious and cultural significance of the black dog, which is regarded as a companion or even as the personification of the devil himself (Dobrovol’skij & Piirainen [2009: 35]). However, Dobrovol’skij & Piirainen [1996: 205] state that the type of animal is not as important for the symbolic meaning as the colour, because the dog is generally an ambivalent cultural symbol, as it does not have exclusively negative connotations, whereas the black colour is clearly associated with magic, devils and demons; and furthermore, it is not decisive for the use of the idiom that the speaker knows about the deeper cultural meaning of the dog, because only the reference to the work ensures that the motivation is understood (Dobrovol’skij & Piirainen [2009: 35]).

52In the scene of Goethe’s book, the devil Mephistopheles shapeshifts from his appearance of a black poodle to his human form, thus revealing himself to the protagonist Faust. The underlying figurative structure can be seen as the understanding is seeing (inside) metaphor.

Table 3. Productivity and variability of X [X= verb] des Pudels Kern in deTenTen20

Table 4. Top 20 types and token frequency of the slot filler

Types X	Token frequency	Percentage
sein	454	61.97%
liegen	98	12.89%
treffen	81	10.66%
haben	28	3.68%
stecken	9	1.18%
kommen	8	1.05%
werden	7	0.92%
dürfen	6	0.79%
scheinen	5	0.66%
bringen	4	0.53%
sitzen	3	0.39%
zeigen	3	0.39%
geben	2	0.26%
können	2	0.26%
denken	2	0.26%
finden	2	0.26%
offenbaren	1	0.13%
benennen	1	0.13%
heißen	1	0.13%
suchen	1	0.13%

53The random example of X des Pudels Kern is an intertextual reference because it is a part of a story called Die Gartenlaube by Levin Schücking published in 1869.

(c) “[…] Herr von Eschebach wollte nach Deutschland zurückkehren – aber da stand bereits Seine Excellenz der allmächtige Minister wieder, streckte seine Hand aus und führte die schöne Wittwe heim.” “Aha – da kommt des Pudels Kern!” rief der Minister mit hohlem, klanglosen Auflachen. “Mein Glücksstern hat den scheelen Neid, die im Finstern schleichende Bosheit gegen mich herausgefordert!” “Excellenz, sagen Sie lieber, die Entrüstung darüber, daß das Böse so viele Jahre lang trinmphiren durfte!” sagte der Portugiese mit starkem Nachdruck und sprühenden Augen. […] (deTenTen20, 12652572797)

Da kommt des Pudels Kern (*there comes the poodle’s core) is used here to directly refer to a person.

4.3. Nel mezzo del cammin di nostra vita

54Nel mezzo del cammin di nostra vita is the initial endecasillabo of Dante’s Divina Commedia (= DC), followed by the verses mi ritrovai per una selva oscura / ché la diritta via era smarrita, which altogether form the first canto. The syntagm, which extends over the first two verses, initially fulfils a deictic function at the narrative level, as the protagonist is placed in relation to space and time. The prepositional phrase of the first verse and the metaphor of the path of life it contains serve as a temporal determination. Following the Bible, the middle of the journey of life is synonymous with the 35^th year of life (Gmelin [1966: 26]). At the same time, the metaphor of the journey through life symbolises the plot of the DC as a whole, as the protagonist who is also the narrator embarks on his personal adventure. The three opening verses are interwoven in terms of imagery and have several intertextual references. In addition to the action dimension of the 35-year-old first-person narrator getting lost in the dark forest, a meta-level is opened up via the opposition nostra vita and mi ritrovai. As the possessive pronoun in the first-person plural emphasises, it is not just about the protagonist himself, but rather about the fate of all people. Giannantonio [1986: 4] states the narrator’s individual life journey traces a universal conditio humana. This interpretation seems plausible, because la diritta via also makes a reference to the Bible and is to be understood as an allegory of the virtuous path of life (Gmelin [1966: 28]). With Aristotle, Thomas Aquinas and Albertus Magnus, further intertextual references can be identified for the first verse, which, like Dante, interpret the path of life as an arco, the climax of which can be quantified as the mid-thirties in reference to the age of Jesus Christ (Gmelin [1966: 26]). Nel mezzo del cammin di nostra vita is therefore metaphorically and intertextually motivated. The metaphoricity is limited to the component cammin, because the spatial concept of the path or route (= it. cammino), constituted by the starting and end point, is projected onto life, the beginning of which is marked by birth and the end by death. The winged word is semantically divisible, or compositional, because only the component cammin is idiomatic. Regarding aesthetic features, it should be noted that the lexeme cammino was apocopated to maintain the metre. It is likely that this feature will be retained in modifications in order to preserve its recognisability and establish a link to the original.

55Syntactically, the first verse consists of the temporal adverbial Nel mezzo del cammin di nostra vita, which functions as a supplement to the main clause, which in turn is composed of a reflexive transitive verb (mi ritrovai) and complement (per una selva oscura). For the second verse, this results in the classic theme-rheme structure with the final focalisation of selva. Paraphrasing the first verse as ‘when I reached midlife’, it can also be interpreted as a rheme in which the subject of the second verse (the subject in the 1st person singular) has already been anticipated by the pronoun nostra. The possessive pronoun can be interpreted here as totum pro parte.

Table 5. Productivity and variability of Nel mezzo del cammin di nostra X in itTenTen20

Table 6. Top 20 types and token frequency of the slot filler X in Nel mezzo del cammin di nostra X

Types X	Token frequency	Percentage
vita	780	96.77%
scuola	3	0.37%
via	3	0.37%
metro	2	0.25%
gita	2	0.25%
vitc	1	0.12%
guida	1	0.12%
story	1	0.12%
udienza	1	0.12%
e	1	0.12%
lettura	1	0.12%
patria	1	0.12%
regular	1	0.12%
vita1	1	0.12%
verita	1	0.12%
x	1	0.12%
serie	1	0.12%
….	1	0.12%
vitami	1	0.12%
vitb	1	0.12%

56The following example found in the corpus is a title of an online blog about mafia-related police investigations and legal lawsuits in Sicily in the 90s.

(d) Nel mezzo del cammin di nostra udienza... Dopo aver sviscerato nei dettagli le contraddizioni di Mannoia e la vicenda dei suoi due interrogatori nascosti alla Difesa di Bruno Contrada, occorre fare un cenno anche alle spiegazioni che i PM Ingroia e Morvillo fornirono circa questa misteriosa vicenda. L’udienza è sempre quella. (itTenTen20, 10397180603)

57The slot filler udienza (audience) disturbs the rhyme scheme of the canonical form but keeps the relation to a period of time. However, there are two slot fillers (metro and gita (en. trip, excursion, jaunt), see Table 6) that can be viewed as remotivation of the literal sense: cammin is semantically re-concretised by referring to distances.

4.4. An eye for an eye, a tooth for a tooth

Table 7. Productivity and variability of a[n] X for a[n] X and a[n] X for a[n] X a[n] Y for a[n] Y in enTenTen21

Table 8. Top 20 types and token frequency of the slot filler for X in a(n) X for a(n) X

Types X	Token frequency	Percentage
eye	5367	49.09%
tooth	2108	19.28%
life	582	5.32%
lot	76	0.70%
favor	61	0.56%
day	59	0.54%
hand	43	0.39%
ear	42	0.38%
plan	41	0.38%
man	33	0.30%
nose	33	0.30%
woman	32	0.29%
soul	31	0.28%
friend	28	0.26%
foot	25	0.23%
blow	23	0.21%
bit	23	0.21%
hour	22	0.20%
gift	22	0.20%
book	21	0.19%

(e) “Exactly. You’d like to murder me, Cousin Ned, to stop my mouth, but I’ll not give you the chance. And, really, we ought not to kill one another, for the Cragg motto is ‘a Cragg for a Cragg.’ That has probably influenced my poor father more than anything else in his dealings with you. He knew you are a Cragg.” “Because the family motto was first ignored by yourself.” (enTenTen21, 25232669432)

59The example above refers to a 1910s series of novels called Bluebird Books, namely the novel Mary Louise in the Country. At first glance, one could think the example is an instance of a(n) X for a(n) X because it has been semantically primed with the verbs to murder and to kill in the preceding sentences. In the context of the novel, however, a Cragg for a Cragg (a Cragg is a member of the Cragg family) does simply mean ‘Craggs support each other’. This shows that the structural template a(n) X for a(n) X is not always used with the same meaning, and it points to the general problem that only case by case analysis can shed light on the quality of the expression.

Table 9. Top 20 types and token frequency of the slot filler for X and Y in a(n) X for a(n) X a(n) Y for a(n) Y

Types X & Y	Token frequency	Percentage
eye, tooth	989	66.60%
life, eye	44	2.96%
tooth, life	34	2.29%
tooth, eye	33	2.22%
eye, life	30	2.02%
eye, nose	23	1.55%
tooth, hand	22	1.48%
hand, foot	19	1.28%
ear, tooth	18	1.21%
nose, ear	18	1.21%
burn, wound	8	0.54%
wound, bruise	7	0.47
strategy, strategy	6	0.40%
eye, eye	6	0.40%
foot, burn	6	0.40%
eye, hand	5	0.34%
life, tooth	4	0.27%
eye, ear	4	0.27%
hand, leg	3	0.20%
tooth, blow	3	0.20%

60The following example for the pattern a(n) X for a(n) X, a(n) Y for a(n) Y is a special case because the two-part phraseme evolves into a three-part one:

(f) […] The only rule is balance. There must always be balance. A question for a question, a feeling for a feeling , a fear for a fear ... and always centered, always addressing the Listener. […] (enTenTen21, 59676865971)

61The author uses the structure a X for a X, a Y for a Y, a Z for a Z to exemplify his notion of balance which corresponds to the idea of equivalence and reciprocity; it thus shares a substantial semantic amount of both the canonical form an eye for an eye (equivalence in punishment) and the modern adaptation (equivalence in retaliation).

4.5. Corpus results

62As can be concluded from the results based on the Hamletian winged word to be or not to be, that is the question the phraseme generated a pattern, however, the structural formula to X or not to X had been conventionalised before. The difference lies in usage: the Shakespearean pattern is loaded and stresses the importance of a decision, whereas it can be used in situations that are not important at all. The results support that claim, since some slot fillers are everyday action verbs like buy, eat, use, go, pay. Furthermore, different levels of structural changes can be identified:

• canonical form: To be or not to be, that is the question

• modification of canonical form (left context): To X or not to X, that is the question

• modification of canonical form (right context): To be or not to be, X

• pattern: To X or not to X

63The results correspond to Pham’s view on intertextuality as gradual, as we move downwards from canonical form, modification of the first part with original right context or original first part with modified right context, to modified fragment / pattern, which is hardly distinguishable from open-choice principle. We can assume that syntactic and semantic processing supports the inference of meaning, and it can be hypothesised that it thus helps identifying intertextuality. Additionally, all the forms can be “tagged” with metacommunicative markers and / or formal cues like quotation marks or other signs.

64The analysed winged words evolved into prototypical expressions. They are constituted by high utility with regards to other possible expressions, in the case of to X or not to X it serves either as elaborate alternative to a generic disjunction (e.g. with or / either… or…) and to load a topic with significance. The winged word an eye for an eye (a tooth for a tooth) is highly variable, less than to be or not to be (that is the question), but it comes close to the same magnitude order. German des Pudels Kern (sein) and Italian nel mezzo del cammin di nostra vita do not tend to variability. Even if the German winged word scores about ten times as many modifications as the Italian one, the slot fillers indicate that it is stable and used in the same way as with the verb sein. The German and the Italian examples are lacking schematicity; they are both syntactically and semantically restricted. However, modifications of other elements of the winged words have not been searched. Exorbitant degrees of propagation seem to be reached with more flexibility, as proven by the Hamletian soliloquy. It can be hypothesised that an apt winged word does both: on the one hand its original form is widespread and used outside of its source domain, and on the other it allows variability for modifications or even evolution into a pattern.

5. Conclusions

65As banal as it may seem, it is no coincidence that winged words often originate in famous stories of renowned authors, which represent not only specific linguistic communities but also structures, networks and hierarchies of essential values that are coherent, culturally rooted and passed on to succeeding generations.

66When Lakoff & Johnson [1980] formulated the idea of the metaphorical nature of the human conceptual system as a starting point for their theory of metaphors as constitutive elements of human cognition, perception and language, alongside with Langacker’s Cognitive Grammar [1987] and focus on semantics, they posed a revolutionary opposition to Chomskyan generative grammar. Especially with regards to orientational metaphors (up-down, in-out, front-back, on-off, deep-shallow, central-peripheral) Lakoff & Johnson [1980: 14-21] conceived cognition in dynamic relationship with perception of and bodily experience in the physical environment that has a huge impact on language. Yet people are not only “metaphorical beings”. “Understanding language means relating language to one’s own actions, possibly because the automatic and extremely rapid linkage of sensory and motor information in our brains benefits comprehension and learning processes” (Pulvermüller [2002: 581]). Humans are highly goal and meaning oriented, in the way they act and in the way they communicate.

67The huge advantage of phraseology is that it is somewhat compatible with a multitude of different scientific disciplines, without running the risk of hitting a dead end by highlighting only one aspect of language as the fundamental one. On the contrary, phraseology has proven its interdisciplinary nature along the lexico-grammatical continuum with regards to usage-based, constructivist, pragmatic, sociolinguistic and corpus linguistic approaches, as well as to memetics, studies on culturemes and formulaic language.

69The nature of the modification of winged words is an aspect that needs further investigation. In the context of this paper, this would mean a qualitative analysis of the results on a case-by-case basis, sufficient to distinguish between occasional modification and semi-schematic phrasemes with lexically open slots, and additionally to identify whether different meanings are assigned to the same structural template. Mellado Blanco [2024: 497] acknowledges the continuum of occasional modifications of phrasemes and semi-schematic constructions, but calls the division based on the number of occurrences proposed by Stumpf [2016] arbitrary. This does not seem to be an issue from a CxG perspective because “variation is approached holistically […] and the creative play of idioms, quotations, proverbs, and so on, is not interpreted as an irregular phenomenon, but simply as infrequent instantiations of constructions” (Mellado Blanco [2024: 507]).

70What if using winged words and patterns and modifications is not a sign of creative laziness? What if it is instead a cost-efficient way of referring to common knowledge, respectively, of recycling famous prefabs to promote new contents? What if these hybrid forms fit our brain structure, because they allow top-down inferences that fit our experiences and expectations but also catch our attention containing new elements to be processed bottom-up?

71If so, wouldn’t studying winged words including the overt knowledge and connected values be the perfect stepping stone to learn both a language and a culture from a learner’s perspective?