‘Rare but there’: On subtleties and saliencies of (pseudo-)archaisms in Game of Thrones scripted dialogues

Abstract

This paper is going to carve out to what extent the use of morpho-syntactical as well as lexico-conceptual archaisms is frequent or salient in the Game of Thrones series dialogue transcripts and what kind of (pseudo-)archaisms the screenplay writers were working with in order to add a ‘pseudo-medieval feel’ to the discourse on screen. By making use of the corpus tool Sketch Engine, I shall demonstrate that archaisms and particularly pseudo-archaisms in Game of Thrones may not be particularly frequent token-wise, but are highly salient, despite their presence being subtle and impalpable. Their function as pseudo-medieval landmarks, however, will be shown to be cohesive throughout the scripted dialogues in the Game of Thrones series.

1 Introduction: Object of this Study

Those who have watched the TV series Game of Thrones (2011–2019) in its original English version will certainly agree that many of the dialogues between the characters do sound slightly more archaic than conversations in the usual given contemporary TV series. At the same time, hardly anyone would perceive the scripted discourse in Game of Thrones as ‘odd’ or disturbing but arguably as consistently apt for what the viewer will conventionally be expecting from a fantasy-medieval TV series: that is, a coherently pseudo-authentic multimodal text within the given pseudo-medieval and fantastic frame. Examples (1)–(6) may demonstrate this:

These samples include subtly inserted archaisms such as address formulae (e.g. dear brother, my lord, my Queen) and phrasings that would be perceived as highly mock-polite today (e.g. You have much to teach me, forgive me); archaic lexical items that would usually be expressed differently nowadays (e.g. whorehouse); pseudo-archaic words that are either not attested for the Middle Ages (e.g. cut throat) or not attested at all (e.g. direwolves); and concepts that are pseudo-medieval (e.g. carrier ravens instead of carrier pidgeons), obsolete or simply non-relatable for a modern viewership (e.g. bastard sons, fighting tournaments, times of the month referred to based on the position of the moon, fathers marrying off their children).

This paper is going to carve out to what extent the use of morpho-syntactical as well as lexico-conceptual archaisms is frequent or salient in the Game of Thrones (GoT) series dialogue transcripts and what kind of pseudo-archaisms the screenplay writers were working with in order to make the discourse sound ‘pseudo-medieval’. A number of terminological and conceptual issues will be discussed before we shall actually end up talking about samples from GoT: Section 2 contains a few preliminary remarks on the genre categories ‘pseudo-medieval’ and ‘fantastic pseudo-medieval’ in order to provide a frame for classifying GoT within the latter genre. Section 3 then attends to pseudo-medieval speech in telecinematic discourse and the two most obvious options at a scriptwriter’s disposal when seeking to add a ‘medieval feel’ to scripted discourse, i.e. code-switching and the (un)systematic insertion of (pseudo-)archaisms. Section 4, preceding the conclusions, forms the main part of this chapter, providing a mixed-method analysis of the full GoT transcripts covering 8 seasons and 73 episodes with regard to specific types of (pseudo-)archaisms woven into the diegetic discourse.

2 The (Pseudo-)Medieval and the Fantastic

First, let us first briefly look at the genre of what has often been called ‘medieval film’ and the phenomenon of ‘New Medievalism’. This is necessary in order to provide the appropriate frame for the upcoming deliberations and arguments on pseudo-medievalisms and pseudo-archaisms in GoT, which I will assign to the genre of pseudo-medieval fantasy.

Merely since the beginnings of talking films, medieval motifs and narratives have been used and adapted for the audiences of modern times. Amongst the earliest English-speaking medieval films were Reilley Raine et al.’s The Adventures of Robin Hood (USA, 1938), Willingham and Wasserman’s The Vikings (USA, 1958) and Anhalt’s Becket (UK, 1964), which may be vaguely categorized as historical romance, historical action and historical drama respectively.

Most research on medieval films published has been contributed by historians and film historians (cf., for example, Harty 1987; Aberth 2003; Amy de la Bretèque 2004; Kiening und Adolf 2006; Bildhauer 2011). The notion of medieval film has traditionally been applied broadly in the sense of “film about or concerning the Middle Ages” (cf. e.g. Kirner-Ludwig 2020, p. 224; 2018; Bildhauer 2011), thus forming a subgenre to historical film (cf., for example, Rosenstone 2006). According to Burt (2007, p. 219), the portrayal of the Middle Ages, or what he calls “movie medievalism”, covers three main strands, i.e. (a) films that are situated in the Middle Ages and anchored to certain historically attested events, (b) films that play in the modern viewer’s time, but jump back and forth, and (c) so-called neo-medieval films that go beyond the pseudo-authentic medieval and include fantastic elements. The latter includes both films and TV series from the genre I will be referring to as ‘pseudo-medieval fantasy’. Note that these labels (as well as many others) have so far primarily been applied to film, i.e. movies, not, however, to TV series, but may just as well be transferred respectively. I will here use the collective abbreviation FTV (F—film, TV—TV series) to cover phenomena occurring in either.

As shown in Fig. 1, I thus, retain Burt’s (a) and (c) Groups as represented by the pseudo-medieval FTVs on the one end and the medieval fantasy FTVs on the other as the two extremes on a continuum. At the same time I will be acknowledging that candidates will be fluidly rather than absolutely positioned on this very continuum, always depending on the focus the makers of that series chose. Note that I discard Burt’s Group (b) altogether, as FTVs that would fall into that category—e.g. A Knight in Camelot (1998) or Black Knight (2001)—will still make reference to historical events or include fantastic features within the medieval setting depicted.

Fig. 1

Source © Monika Kirner-Ludwig

A continuum of FTVs with examples.

I personally find the notion of ‘medieval film’ (or series) rather misleading, given that anything depicted in FTVs concerned with the Middle Ages can be nothing but inadequate: it will necessarily be based on stereotypical mesh-up ideas brought forth by what are usually non-experts of Medieval Studies. I therefore add the prefix pseudo- to medieval (cf. Fig. 1 and elsewhere) not so as to criticize the quality of films concerned with the Middle Ages in any way, but merely so as to acknowledge their (willful) historical inadequacy and their intent to ‘play out’ some version of Middle Ages for their audience. Medieval fantasy FTVs, on the other hand, do not have to pretend. The genre allows the viewer to expect anything and everything, the impossible, the magical, the incomprehensible, with stereotypically medieval components being but part of that potential range. The latter is also what e.g. Richard Utz (President of Studies in Medievalism) called “neomedieval”, rightly observing that

[n]eomedieval texts no longer need to strive for the authenticity of original manuscripts, castles, or cathedrals, but create pseudo-medieval worlds that playfully obliterate history and historical accuracy and replace history-based narratives with simulacra of the medieval, employing images that are neither an original nor the copy of an original, but altogether ‘neo.’ (Utz 2011)

Thus, GoT is grouped within that latter category, together with TV series such as Merlin (2008–2012) and The Witcher (2019–2021), i.e. two series that, just like GoT, do not have any historio-authentic reference points but are fully set in worlds where magic is a reality and where stereotypically medieval features such as swords, knights, castles etc. provide the sketchy conceptual frame.

3 Pseudo-Medieval Speech in Telecinematic Discourse

Before we attend to the question of how the telecinematic discourse (TCD) in GoT is layered and composed so to actually convey pseudo-medieval authenticity, let us take a step back and consider the complexities entailed in making the diegetic dialogue in a TV plot sound ‘medieval’ to the audience without actually disrupting the viewer’s smooth processing of what they see on screen. Pseudo-medieval or not, the base metalanguage, i.e. the language in which the baseline story is told (let’s say English) must usually be coherent, easy to follow, and has to be appropriately fitting for what are scripted but seemingly spontaneous dialogues on screen.¹ At the same time, the speech used amongst the characters must also be perceived as authentic within the contextual frames given.

When it comes to pseudo-medieval and fantasy-medieval FTVs, I propose that filmmakers and scriptwriters have two salient options at their disposal to make a text ‘sound medieval’: for one, they may weave in (pseudo-)archaic components into their scripts, i.e. archaically cliché building blocks such as mylady, Sire, thou, ye, etc. (cf. Sect. 3.2). For another, they may have their characters code-switch between what is usually the baseline or narrative language (i.e. English) on the one hand and pseudo-medieval or even constructed (i.e. fictional) varieties on the other hand (cf. Sect. 3.1).

Let me forestall that—as I have argued elsewhere (2020)—in pseudo-medieval as well as medieval-fantasy TCD neither one of these two options is encountered particularly frequently. At the end of the day, most pseudo-medieval films since the 1960ies have been found to have used both strategies in a rather unsystematic manner, which affirms that it will mostly be the visuals and musical makeup framing the TCD that is deemed to generally suffice to make the audience perceive the multimodal text on screen as ‘medieval’ and pseudo-authentic.

3.1 Code-Switching

In the past 20 years and certainly in lockstep with the rise of Tolkienian fandom, it seems that the desire for more linguistic authenticity in pseudo-medieval and fantasy-medieval FTVs has been increasing—not only amongst historio-linguistic scholars but in fact amongst lay-viewers highly fascinated by constructed languages such as Sindarin, Quenya just as much as Old Norse and Old English (OE), the latter two of which have recently been scripted into The Vikings (2013–2020).² Thus, while the average viewer will probably not (be expected to) mind grammatically correct or incorrect sequences in pseudo-medieval varieties, but usually be satisfied with whatever sounds ancient and mysterious, there seems to be a rising trend among filmmakers to seek advice from linguists with regard to pseudo- or fictionally linguistic sequences so as to at least avoid major slips. The upcoming examples show code-switching between the base language English and pseudo-Old English (7), Latin (8), as well as two constructed languages featuring in GoT, i.e. Dothraki (9) and Valyrian (10). Particularly the Old English excerpt shows that the character Athelstan makes two major mistakes, i.e. “haþeþ” instead of OE hafaþ (3rd person singular indicative active of habban ‘have’) and “jure” as a result of interference with Modern English your as the second person singular possessive determiner instead of thine. Whether or not these mistakes were scripted as such cannot be determined, but the fact that westseaxe is used in reference to the land the Vikings have arrived on (i.e. Wessex) is an anachronism: the bare compound is only attested as a reference term to the demographic group of West Saxons, not however to the region itself (cf. OED). The correct phrase would have been Westseaxena rice.

No matter if screenplay writers choose to go with a pseudo-medieval variety as they did in (7) and (8),³ or with constructed languages as in (9) and (10), the main function that such varieties fulfil is to convey a pseudo-medieval, foreign feel—with characters speaking languages that do or at least could stem from another time and age and with viewers not being supposed to actually understand what is being spoken (about).

Examples featuring Latin (as in (8)) are usually constructed grammatically correct and usually pronounced neatly (while being accompanied by subtitles anyway), thereby catering towards Latinophile viewers in particular.⁴ However, (11) is a case contradicting to this general impression: Again, we cannot know whether the script itself had been grammatically correct and whether it was the actor remembering the Latin wrong, but the released sequence contains several mistakes on the syntactical, the morphological and the idiomatic level: sanguinarius, which is supposed to translate ‘bloody’ in a metaphorical, cursing sense is only attested as ‘bloodthirsty’ in Latin; the utterance “ego numquam pronunciare mendacium” lacks an inflected verb form in the present tense indicative, cohesive with the first person singular (i.e. pronuncio); and the pronoun ego ‘I’ would not usually have been used to begin with unless specific emphasis would have been intended, given that the inflected verb form sum ‘I am’ entails the personal reference. In example (8) above, however, the switches are grammatically correct and even sophisticated enough to contain a correct hortative form in the present conjunctive (i.e. habeamus ‘let us have’).

3.2 Archaisms and Pseudo-Archaisms

While archaisms may be defined as “linguistic forms that used to be common but then went out of fashion” (Traxel 2012, p. 42), pseudo-archaisms are such “linguistic forms that never existed but […] evoke the impression as if they could have” (Traxel 2012, p. 42 f.). Needless to say that the average viewer will not be expected to recognize the difference between these two phenomena, but also the screenplay writers cannot be assumed to make any such choices deliberately, given that they will usually lack any expert knowledge about the historical stages of English. That is even true for George R. R. Martin himself. His and most other screenplay writers’ (quasi-)knowledge about the Middle Ages (not to even mention medieval varieties) will hardly exceed beyond what their audience believes to know about the Middle Ages. Neither the pseudo-medieval nor the medieval-fantasy FTV-maker will therefore usually tread upon the shaky grounds of incorporating medieval varieties to begin with, as it might generally not be considered worth the effort. After all, for the general viewership, a few pseudo-archaic chunks sprinkled here and there will arguably be sufficient to produce the desired effect: i.e. that of language sustaining the main channel of meaning, which is the visual and auditive one.

This being said, Traxel proposes to subdivide pseudo-archaisms into mock-archaisms, i.e. forms that are composed on the basis of “no or only limited knowledge of English language history [and] created mostly for humorous reasons” (Traxel 2012, p. 43) on the one hand and such formations that are created as neo-varieties (e.g. neo-Old or neo-Middle English) “by authors with an educated knowledge of English language history” on the other (Traxel 2012, p. 43; cf., for example, Görlach 1981; Lenard and Walker 1991). These latter are rarely found overall—in Tolkien’s works for instance as well as in few movies that have been investigated to that end. Traxel’s distinction gets shaky when it comes to cases such as Mel Brook’s Robin Hood: Men in Tights (RHMT 1993), which has been shown to feature numerous mock-archaisms that are indeed very much based on quite an understanding of e.g. Latin morphology and medieval English syntax, morphology, lexis and phonology (Kirner-Ludwig 2020), cf. e.g. (12) and (13). This being said, I will not make further use of the notion ‘mock-archaism’ here, as it seems inappropriately restrictive, but shall rather speak of pseudo-archaisms as opposed to archaisms.

Examples like (12)–(22) were discussed in a study published in 2020, where I proposed a list of (pseudo-)archaic options that I identified in 12 pseudo-medieval and fantasy-medieval movies released between 1963 and 2015. That list (cf. Table 1) shall serve as my springboard into the upcoming analysis of the full episode transcripts of GoT.

Table 1 Ranked list of (pseudo-)archaic features in a corpus of 12 pseudo-medieval/fantasy-medieval movies (1963–2015). Source © Monika Kirner-Ludwig, table adapted from Kirner-Ludwig (2020)

4 Data and Method

This study takes into consideration a corpus including all 73 episodes in all released 8 seasons of GoT (2011–2019). The dialogue transcripts were all obtained from https://gameofthronesfanon.fandom.com/wiki/Category:Transcripts in July of 2019, extracted into txt documents and compiled as a corpus of all in all 506,540 words. This count yields an average of 63,318 words per season, an average of 6,939 words per episode, and an average count of 867 sentences per episode.

It should be noted that I adhere to post-production transcripts of telecinematic discourse only, i.e. not the original screenplays for the series, as the latter are simply not available publicly. The transcripts here used have been produced by fans of the series and provided online for anyone like-minded. Making use of post-factum transcripts rather than the original screenplays certainly presents a few caveats, considering that they are likely to deviate from whatever the original script used to be for several reasons: for one, actors working from the original script may already have embedded idiosyncratic features into their ‘delivery’ of the written screenplay text, including, for example, prosodic and paralinguistic features or even particular accents of English. They may even have replaced certain lexical items by their preferred choices, which means that the upcoming discussions concerning vocabulary will need to take such possibilities into account. Thus, at the post-production stage, i.e. when viewers get to watch the series being broadcast, the originally scripted dialogues may already have been altered considerably. It is then that ‘lay’ transcribers (i.e. fans with presumably no linguistic training) will transcribe the spoken discourse presented on screen back into written form. However, to what extent the individuals taking on this task will do so reliably and diligently may vary extensively, too. As shall be shown, some will even insert extradiegetic information contextualizing the dialogue, which I shall be referring to as pseudo-stage directions (PSDs) in the following. PSDs account for about 2,000 words in the corpus under investigation.

The transcripts were uploaded, compiled and analyzed using the corpus tool Sketch Engine, which provides word lists, concordances and keywords in context (KWICs). I use these as my springboard into a mixed method study on the frequency and saliency of certain (pseudo-)archaisms in the scripted dialogues in GoT. The phenomena that I am going to focus on are restricted to three layers of pseudo-archaic language, i.e. modal verbs, word formation and lexico-semantic choices, all of which ranked amongst the three features most frequent in Table 1. Furthermore, my selection is based on a preceding piloting stage in which all transcripts were semi-automatically and manually browsed for respective examples. In the course of that I was able to already exclude such phenomena that were not present at all in the corpus, including any (mis-)use of archaic personal pronouns or determiners, archaic syntax,⁵ (pseudo-)archaic inflection, contractions, salient use of homonymy and anachronistic lexico-semantic choices.

5 Saliency versus Frequency, or: Dissecting the Discursive Layers in Search for (Pseudo-)Archaisms in GoT

In the following I will argue and demonstrate that archaisms and particularly pseudo-archaisms may not be particularly frequent token-wise, are however highly salient in GoT.

5.1 Modal Verbs, or: The Archaic Use of Shall

The modal verb shall occurs 134 times in the corpus, indicating a future reference that would usually be expressed by the use of will-future or going to-future in Modern English. In contrast, the modal verb will occurs 3,586 times in the corpus. Thus, is not frequent but salient (even though it is not listed as significantly key in the corpus), given that it additionally entails the implication of inevitability, personal determination or compulsion of fate, as the following examples show. Note that shall as a future marker was only replaced by will as recently as the 17th century.

5.2 Archaic Word-Formation

The GoT corpus contains a range of (pseudo-)archaic formations that do not stick out by frequency at all. For one, compound adjectives containing {born} as their endocentric head occur in a number of variants in GoT (cf. Table 2).

Particularly frequent is ironborn, which is concretely used in reference to a seafaring people Iron Islands off the west coast of Westeros, cf. (26), (27).

Table 2 Compounds featuring {born} as a second component. Source © Monika Kirner-Ludwig

Stormborn is Daenerys Targaryen’s cognomen, used when others refer to her, or when she refers to herself in an official manner; cf. (28) and (29).

Other than ironborn or stormborn, highborn is not tied to a specific individual or people, but refers to the social status of e.g. Jon’s father (31) or other noblemen and women. However, in most instances (in 15 out of 22, to be exact), highborn is used in a disparaging manner by someone lower born. The much less frequent adjective trueborn, on the other hand, is consistently used in a positive manner and incomplementarily to bastard. Base-born, which only occurs one single time in the corpus, is a synonym to lowborn, but a pseudo-archaism, considering that it is attested only after the Middle Ages, i.e. since 1553 (cf. (34)).

Further, we encounter 11 mentions of the noun raper (none of rapist) in the corpus. This derivate of the verb rape is first attested in 1878 (OED), so does qualify as a pseudo-archaism—which would have been true, too, for the more common noun rapist, which, too, is first attested for the second half of the 19th century (1869, cf. OED). However, if the writers were aware of these two options, their choice of raper over rapist was an informed one, given that the suffix {ist} as in rapist is in itself rather young (attested from 1842 onwards), whereas the suffix {er} in raper has been attested since Old English.

Finally, a handful of past participle forms of (formerly) strong verbs occur in their archaic forms in GoT, featuring the inflectional suffix {en} instead of the weak dental suffix, e.g. misshapen (misshapen, (35)) and molten (< melt, (36); OED: “chiefly archaic”), downtrodden (< downtread, (37)), beholden (< behold ‘bound to’, (38)) and forlorn (< †forlese, (39)). What is important to mention, though, is that both occurrences of molten, one of the three occurrences of downtrodden and seven out of eight occurrences of forlorn in the corpus do actually not occur in the diegetic dialogue, but in the pseudo-stage directions (cf. e.g. (40), (41)).

5.3 (Pseudo-)Archaic Vocabulary

The discussion of formations in {born} in 5.2. allows for a smooth transition into this final section on (pseudo-)archaic vocabulary in GoT. The lexical layer is by far the most prominent dimension featuring pseudo-archaisms, which was also true for the 12 films investigated in my earlier study (2020).

Of course, the significant prominence of pseudo-medieval vocabulary is per se not surprising, given that the setting of the narrative is fantastic-medieval to begin with. In other words: things and concepts will have to be called by their names or be shown on screen. A number of studies have addressed the most stereotypical concepts typically associated with the Middle Ages and as I have shown elsewhere (2020), pseudo-medieval and fantasy-medieval movies have been relatively consistent in terms of referring to respective topics throughout. For one, these are concepts associated with castles, nobility and armory (cf. also Traxel 2008, p. 130; Sturtevant 2010, p. 5) as well as with romance (cf. Hasty 2016). Both are represented in merely all pseudo-medieval as well as fantasy-medieval FTVs. Sorcery, curses, magic and dragons are similarly salient across the fantasy-medieval genre.

This being said, I investigated the GoT corpus in regard to vocabulary that would be representative of these topics and concepts in order to carve out what the transcripts as a whole are made up of in terms of lexical material. Due to Sketch Engine’s automated parts of speech (POS) tagging function, the user has an efficient time to separate all function words (e.g. prepositions, determiners, conjunctions) from the lexico-semantic material of interest, as the former are not further relevant to this study. Function words make up for ca. 43% of the corpus, which left me with 57% of the word stock to work with. Out of that, only 7.6% responded to the lexico-semantic fields or frames under investigation (cf. Table 3).⁶

Table 3 Stereotypically medieval and fantasy-medieval frames as represented in GoT. Source © Monika Kirner-Ludwig

Let us focus on those small groupings making up less than 8% of the vocabulary. Many of these words are statistically significant in key when compared to a reference corpus, in this case the English Web 2018 (enTenTen18, ‘reference corpus’), which is about 33,600 times larger (21,522,585,853 tokens) than the GoT corpus (cf. Table 4a–4f). The cut-off point for MI scores was 07.000.

Table 4a Kings, queens and servants. Source © Monika Kirner-Ludwig.

Table 4b Weapons, war and violence. Source © Monika Kirner-Ludwig

Note that archaic concepts in relation with weapons, war and violence are also expressed in archaically sounding verb phrases such as readying an arrow, a horse, a saddle, the shield, the sword, etc., although it needs to be highlighted that most of these actually occur in the PSDs rather than the diegetic dialogue; cf. (46)–(50).

Table 4c Family and legacy. Source © Monika Kirner-Ludwig

Table 4d Sex, alcohol and game(s). Source © Monika Kirner-Ludwig

Table 4e Medieval spaces. Source © Monika Kirner-Ludwig

Table 4f The fantastic and the supernatural. Source © Monika Kirner-Ludwig

5.4 (Pseudo-)archaic Address Formulae

The use of (pseudo-)archaic address formulae is dispersed across the complete corpus of GoT and adhered to by almost all characters when addressing anyone higher in rank or in an intentionally polite manner. The distributions are shown in Table 5.

Table 5 (Pseudo-)archaic address formulaeFootnote

The asterisk here and elsewhere indicates a so-called ‘wild card’ search. Browsing for < my princes* > will yield results for my princes, my princess, my princesses, my princes’ etc.

. Source © Monika Kirner-Ludwig

The address formula < determiner + grace > as a term of address of a superior individual has indeed been in use since the 12th century, most commonly with the premodifying 2nd person possessive determiner, i.e. Your grace, which is used to address ladies as well as men of higher rank. The formula also occurs with the 3rd person possessive determiners his and her (OED grace n.).⁸ Interestingly, though, the formula my grace addressing someone of higher rank, which occurs but twice in the full corpus, once addressing Cersei, once addressing Daenerys, is a pseudo-archaism. It is not attested for Middle English or even later than that and seems to be an interference—either on the scriptwriters’ or the actors’ part—with the formulae my lord, my lady etc., which do in fact use the first person singular possessive determiner.

Men and women of higher or similar rank as the lady addressed would usually use my lady or < Lady + first name > ; Lord and my lord is used respectively:

In fact, my lord seems to be used for individuals ranking higher than someone to be addressed with your grace, as the viewer learns in S2 E3, when Catelyn addresses King Renly Baratheon and is criticized by Brienne. A few turns back, Ser Colen of Greenpools, a knight, addresses the king as your grace (60).

The archaic and well-attested formula your lordship only occurs three times in the GoT corpus, twice used by Bronn to address Tyrion and once by Shae addressing Tywin. Why this particular formula was inserted so rarely and unsystematically is unclear—in other instances, Bronn uses my lord to address Tyrion as well as other higher ranking individuals. Interestingly, the differentiation between m’lord and my lord is addressed by Tywin in S2 E7; cf. (67).

The address titles ser/sir are used interchangeably in the GoT corpus and cannot be clearly distinguished based on the transcripts (given that they are quasi-homophones). All in all, together they occur 598 times, with 24 male individuals being addressed once or several times. Out of these, 17 are attestedly knights (cf. https://gameofthronesfanon.fandom.com/wiki/, last accessed 3 September 2021),⁹ 7 probably are but cannot be verified as such,¹⁰ and Ser Jorah Mormont used to be one.¹¹ It should be noted that ser is a rarely attested spelling variant of sir and sire (with sire occurring only three times in the corpus) and is now explicitly labelled as an archaic title of distinction, as opposed to sir, which is still very much in conventional use (OED). Examples (74) and (75) show ser in use during the 15th century.

5.5 Recurrent Archaic and Obsolete Concepts

Archaic and partly obsolete themes are subtly recurrent in GoT, with one of these being the concept of the oath as a “solemn or formal declaration invoking God (or a god, or other object of reverence) as witness to the truth of a statement, or to the binding nature of a promise or undertaking”(OED), e.g. made by a liege to their king. Occurrences are not frequent, do, however, show a rather wide range of collocates (cf. Table 6) and appear throughout the series with the exception of Season 8, as displayed in Fig. 2.

Fig. 2

Source © Monika Kirner-Ludwig

Occurrences of oath across the 8 seasons of GoT.

Table 6 Collocates of oath in GoT. Source © Monika Kirner-Ludwig

Shame is another concept that is archaically colored in GoT, being closely associated with or even resulting from dishonor or disgrace. As for the noun shame, only 29 out of the 45 occurrences seem in fact relevant for this argument, as in 16 instances shame occurs in the phrase (such/what a) shame, which is void of the emotional weight that shame carries in its bare, non-idiomatic use (cf. Table 7). Particularly the fact that shame is publicly paraded and judged is a recurrent theme in GoT.

Table 7 The concept of shame. Source © Monika Kirner-Ludwig

Throughout the series, women—with the obvious exceptions being Arya, Daenerys and Missandei—give in to the rather traditional and old-fashioned role of the obedient wife (cf. e.g. (3) above), the caring mother and the passive object of sexual desire. This is reflected in some of the archaic language items (apart from numerous utterances that express respective role distributions without making use of any archaic vocabulary). For one, transitive verb phrases such as to bed (a female) someone in the sense of ‘taking a woman to bed for intercourse’ occurs 9 times in GoT. It is labelled as archaic by OED (cf. bed n., 3.). For another, we find the zero-derived verb whore in 8 occurrences, one of which is a pseudo-archaism featuring it as a reflexive verb (cf. (85)).

This being said, certain vocabulary items are restricted to males—no single woman in GoT beds anyone. This is also true for certain lower style items, such as the verb phrases piss and take a piss, which occur 23 times and only 2 instances are uttered by women breaking with that ‘rule’ (even though in (86) Arya is merely repeating Gendry’s choice of words).¹²

The adverb aye in GoT is also clearly a word of masculine, rough style, which is evident from the fact that all ayes but two are uttered by men across lower ranks (e.g. (88), (89))—it occurs 137 times in the corpus. Only Yara and Ygritte get to use this affirmative adverb occasionally, which supports the impression of Ygritte as a wildling and of Yara as a woman of a relatively masculine type (cf. (90)–(92)). Note that aye borders between representing a pseudo-archaism and an archaism, as it is attested only since 1576 (OED).

A few archaic adverbials and prepositional phrases also appear in the corpus (cf. Table 8). Examples (93) to (97) show them in context.

Table 8 Archaic adverbials and prepositional phrases. Source © Monika Kirner-Ludwig

6 Conclusion

As my corpus analyses have shown, language seems to play a comparably marginal role frequency-wise in contributing to the intradiegetic scripted TCD sounding ‘medieval’. GoT does not overdo the use of pseudo-archaisms or, e.g., archaic pronouns at all. If anything, it is the subtle layers of vocabulary that bring the pseudo-medieval feel home. No mock-archaisms or any frame-breaking archaic oddities are used. Language is not ridiculed at all, but respectfully woven in so as to convey coherent authenticity within the frame of the diegetic narrative. This is achieved by a diligent and rather complex texture of linguistic layers blending both pseudo- as well as genuinely archaic words and phrasings together.