Introduction

I have argued before for using contemporary television poetics to re-consider features of ancient Greek epic (2016); here I will argue for using well-studied features of ancient epic to consider contemporary television narrative, refine analogies of narrative strategies between the two forms, draw out the limits of such analogies, and point towards areas ripe for further research in the future. My main focus will be on epic’s repetitive forms, from its oft-repeated epithet-noun formulae and proper names, to its formulaic phrases, to its type-scenes, and how these find corollaries in contemporary television narratives. I will also look at how these forms contribute to both forms’ recall strategies, suggesting a new taxonomy for television recall strategies. Finally, I suggest that applying philological and close-reading methods to television, still largely tools for popular critics and so-called affirmational fans (Busse 2013, 83) rather than television scholars, might further elucidate television narrative forms, while continuing to refine these models of repetition can also breathe new life into Homeric scholarship.

Introduction to Oral Epic

The Ancient Greek epics were most likely composed over hundreds of years in the period leading up to the 8^th century BC. Produced through an oral tradition, many different poets probably contributed to their final form, which was most likely written down sometime in the 7^th century BC. The two longest and most canonical epics, attributed to Homer, are the Iliad and the Odyssey, with the Iliad coming in at over 15,000 lines, and the Odyssey coming in just over 12,000. For the Iliad, most scholars put its performance time at between 18 and 25 hours (Kozak 2016, 2): my own performance of the whole poem, in a fluid English translation, took nearly 27 hours.

One of the dominant scholarly debates about the Homeric epics over the past several hundred years was whether or not the Iliad’s fairly unified story was the product of a single mind (the Unitarians) or of several (the Analysts), with most contemporary scholars understanding the poem as emerging from an oral tradition (Oralists), that draws on a wide range of stories from what is known as the Epic Cycle (the Neoanalysts)¹. This leaves us with the Iliad as a very long, non-literate story that stretches out over 20 hours, and is the product of multiple creators drawing on a traditional form: this should sound familiar to anyone who closely follows contemporary North American narrative television shows.

Whether or not the epics that we have today were the product of a single or many minds, real questions remain as to how such a long poem could have been performed in such a way to keep the story straight, both for the poet and for the audience². Unlike television, the Iliad has no show bible, nor do we know of anything approaching a “previously on” segment: but much like television, Homeric epics rely heavily on various forms of repetition to reinforce important narrative information for both the performer and their audience.

Repetition, Repetition, Repetition

Both Michael Newman and Jason Mittell have spoken about television’s propensity to repeat narrative information over and over again; Newman points to the needs of an audience who might step away from the television or who might not tune in every week (Newman 2006, 19; Mittell 2010, 79); Mittell later sees repetitions tied to increased narrative complexity and serialization (2015, 180–87). But unlike those who study Homer, television scholars have not had millennia to create taxonomies of repetition types. It’s worth briefly going through the different kinds of repetition that we find in Homeric epic in order to consider which of these types is helpful in how we understand contemporary scripted North American narrative television.

The Homeric epics are in metre – dactylic hexameter – so that there are a certain number of rhythmic beats in each line. This means that there are various metrical formulae – words or phrases that take up a certain amount of metre – which a poet can rely on, and which become familiar to an audience. These formulae can include names and their epithets, like κρείων Ἀγαμέμνων (“powerful Agamemnon”; 1.102, 1.130, 1.285, 1.355, 1.411), which can serve to reiterate a character’s traits, or role (“the lord of men”, which appears over 100 times in conjunction with Agamemnon in the poem), or family (he’s called “the son of Atreus” over 100 times as well). This reiteration of character details comes very close to what Newman describes as a feature in episodic North American television (Kozak 2016, 7):

(Recapping) takes many forms, one of which is the perpetual naming of characters: in every beat, characters address each other by name, often several times in a two-minute segment. Along with naming comes role reiteration: Alias (2001) constantly reminds us that Jack and Irina are Sydney’s parents; Giles is always reminding Buffy (Buffy the Vampire Slayer (1997)) that he is her watcher; Joel on Northern Exposure (1990) is often called “Dr Fleishman,” even away from his practice, and his favourite leisure activity, golfing, reinforces his role stereotype. (Newman 2006, 18)

The interesting thing is that despite Homeric scholarship’s claims that the repetition of names and epithets in the epics is a fairly unique result of its oral composition, and even “naïve” (Schein 1984, 1), television shows fairly similar frequencies, if not greater: Lucifer’s eponymous hero, played by Tom Ellis, calls Detective Decker, played by Lauren German, “Detective” a whopping 605 times in the first three seasons of the show, about 40 hours’ worth of television (2016). This compares well to the most common proper nouns in the Iliad, though of course these are said by many speakers, with the “Achaians” mentioned 608 times and “Hektor” mentioned 450 times over the Iliad’s approximately 25 hours. If we think of roles, rather than names, like “the Detective”, the phrase “lord of men” only appears 52 times in the Iliad and even the simple “lord” (ἄναξ), applied to multiple characters, appears just 151 times. If we think of “the Detective” more as a place-holder epithet for Chloe’s name, we might compare it to Game of Thrones’s (2011) characters calling Jaime Lannister (Nikolaj Coster-Waldau) “Kingslayer”: that happens only 24 times in the first five seasons or 50 hours of the show, and Game of Thrones (2011) is fairly unique in contemporary North American television series in presenting this kind of epithet (2011). The Iliad’s character-based epithets are much more common and much more frequent, with Hektor being “shiny-helmed” 37 times, and Achilles being “fast-footed” 51 times (πόδας ὠκὺς Ἀχιλλεύς 30 times; ποδάρκης 21). The most common epithet, δῖος, (“brilliant”, “of Zeus”) appears 222 times, applied to a whole host of different characters. Bakker calls an epithet “the quintessential property” of the character it modifies (1997, 161): further exploration might consider how television also uses visual and musical cues in this capacity, from costuming choices to themes.

Formulaic phrases can be found beyond these repetitions of names, epithets, or titles. The rally-cry variations of “be men, friends!” appears 10 times throughout the Iliad (5.529, 6.112, 8.174, 11.287, 15.487, 15.561, 15.734, 16.270, 17.185). This kind of Homeric repeated phrase deviates from North American cartoon and sitcom character catch phrases, from Homer Simpsons’ “D’Oh!” in The Simpsons (1989) to Joey Tribbiani’s “How you doin’?” in Friends (1994), as these formulaic phrases alternate between characters. Yet these phrases still have television analogues, from Game of Thrones’s (2011) “a Lannister always pays his debts” or “winter is coming” to The Wire’s “It’s all in the game” (2002), to Friday Night Lights’s “clear eyes, full Hearts, can’t lose” (2006). Rather than define a single character, this kind of phrase repetition between characters reiterates the rules of the story-world that the characters reside within. With the Iliad, this battle-exhortation shows up more frequently than even “winter is coming”, which appears only 15 times in the nearly 80 hours of Game of Thrones (2011).

Homeric formulaic phrases most frequently describe character action. The epithet-formula for Agamemnon from above often expands out into simple action in the full-line formulae phrase τὸν δ' ἀπαμειβόμενος προσέφη κρείων Ἀγαμέμνων (“but answering him, powerful Agamemnon said…”; 1.285, 1.130, 2.369, 4.188, 10.42). These formulae introducing character speech are by far the most common and might be likened not to anything said in television, but instead to the formulaic, alternating shots between two people in dialogue in most single camera shows. More generally, Homeric epic’s described repeated action finds analogues in television’s shown repeated actions. Take for instance the Iliadic phrases, “thundering, he fell” (δούπησεν δὲ πεσών; 21 occurrences) to describe the deaths of men, or “he fell from his chariot” (ἤριπε δ' ἐξ ὀχέων; 9 occurrences), or “his armour clanged around him” (ἀράβησε δὲ τεύχε' ἐπ' αὐτῷ, 10 occurrences). The general formulae can be compared to any regular shots of those falling in any shows where characters are frequently killed, from The Walking Dead (2010) to Banshee (2013) to Game of Thrones (2011): if we want a more specific repetition in battle sequences, think of the description of Achilles’ “Pelian ash spear” (16.143, 19.390, 20.277, 21.162, 22.133), and how we might compare it to Jon Snow’s references to and uses of his sword, “Longclaw”, in Game of Thrones (18 episodes).

These kinds of scenario-specific repetitions often signal what Homeric scholarship calls “type-scenes”. Battles are not the only kinds of type-scenes, though of course in the Iliad they are the most frequent. Mark Edwards, in his catalogue of Homeric type-scenes, demonstrates just how wide-ranging these scenes can be, ranging from “arming” and “battle speeches” to “divine visits”, “dressing and adornment”, “travel by sea”, “prayers”, “oath-taking”, “consolation” and “funeral rites”. Eating, too, has its place, and it might be no surprise to anyone familiar with the Iliad and the Odyssey that people eat more often in the latter than in the former. The formula for starting to eat, “They threw their hands at the good things that were prepared, lying before them”, appears three times in the Iliad; and ten times in the Odyssey, while the formula to close a meal, “They had put away their desire for food and for drink” appears seven times in the Iliad and fourteen times in the Odyssey. Certain shows also draw attention to the repetition of their characters’ eating habits, from Liv Moore (Rose McIver) eating her next brain in the morgue in iZombie (2015) to the many dinners around Hannibal Lecter’s (Mads Mikkelsen) dining room table in Hannibal (2013).

Just as battle and eating are common in the specific story-worlds of the Iliad and the Odyssey, we might say that each North American television show, too, has its own specific set of type-scenes, scenes which repeat, with variation, throughout the series, in both episodic and serialized shows. These type-scenes essentially frame important features of each show’s story-world, and can be as wide-ranging as the shows themselves, from visits to the therapist (The Sopranos (1999); Hannibal (2013); Lucifer (2016)) to a crime-solver’s visit to the morgue (Pushing Daisies (2007); Homicide: Life On the Street (1993)) or the lab (Lucifer (2016); Bones (2005)) to a school drop-off (Big Little Lies (2017)), a trip to a gender-neutral bathroom (Ally McBeal (1997); Battlestar Galactica (2004)), or a trial scene (The Good Fight (2017); Law & Order (1990)). We can consider the great number of sitcoms that heavily rely on the type-scene of a group of friends hanging out in the lead’s apartment (Seinfeld (1989); Friends (1994); The Big Bang Theory (2007)); think of your favourite television show and I guarantee you can think of a type scene that happens on a regular basis throughout the show, with episodic shows like procedurals and sitcoms probably more likely to rely on such scenes. This repetition in episodic television might take the place of a serialized plot structure to better immerse the viewer in its storyworld.

We can see different levels of specificity within repeated types of scenes. Consider Game of Thrones (2011), and the difference that exists between “small counsel meetings” scenes (EPS) and scenes where “a direwolf protects a Stark child” (EPS). In considering Homeric type scenes, for instance, Edwards insists that “verbal repetition between different instances of a type-scene may or may not occur” and goes on to follow Nagler’s insistence that “there is no ‘standard’ form of a type-scene from which given examples may be said to deviate more or less”. Yet still, specificity matters: consider the difference between a standard battle scene, with its formal elements, and the more specific instances of a god intervening on the battlefield to save a hero, which occurs with a variety of gods, men, and in a variety of battle scenarios. We might also consider the differences between the two Game of Thrones (2011) type scenes (“small council meeting” scenes and scenes where “a direwolf protects a Stark child”) I suggested above. In Game of Thrones (2011), a small council meeting often has the similar formal elements of the chairs around the table, the various roles fulfilled by various peoples, and some conversation about the affairs of the realm: these formal elements are played upon for various affects, comedic or otherwise, as the series goes on, such as Tyrion moving his chair in the third season’s “Walk of Punishment”, and again in the series finale “The Iron Throne”. But can we say the same for the scenes of direwolves protecting Starks? They certainly reiterate a fact of the story-world, but they might not otherwise be thought of as a set of type-scenes in the same way that the small council scenes can be.

Confused About Callbacks

This confusion continues when we think of the type-scene in connection with television’s understood notion of “callbacks”. In my book, I conflated the two things as one in the same (Kozak 2016, 70), and while the two do overlap, it is worthwhile to try to differentiate how we classify these scenes. When thinking of the small council scenes from Game of Thrones (2011), we might understand the small council scenes themselves as type-scenes, but the series finale repetition of Tyrion moving the chair feels more like a callback, purposely designed to recall another specific scene, rather than just to reiterate a type of scene. We can make the same differentiation in the Iliad. Hektor’s and Patroklos’s death scenes both have typical elements about them and fit easily into typical battle scenes (Garland 1981; Bernard 1968). But when Hektor dies, several hours of story-time after he kills Patroklos, the exact same lines are used to describe both men’s deaths (16.855-7=22.361-3), in a clear callback that links the two men’s death specifically:

ὣς ἄρα μιν εἰπόντα τέλος θανάτοιο κάλυψε

ψυχὴ δ' ἐκ ῥεθέων πταμένη Ἄϊδόσδε βεβήκει

ὃν πότμον γοόωσα λιποῦσ' ἀνδροτῆτα καὶ ἥβην.

As he spoke, death’s end covered him,

And his soul went flying from his limbs towards Hades,

Mourning its fate as it left manliness and youth behind.

While this differentiation between a type-scene and a callback feels helpful, it doesn’t tell the whole story, and actually serves to emphasise just how frustratingly elusive a clear taxonomy of callbacks in either form seems to be.

There are no scholarly definitions of television callbacks. Lavender-Smith sees the term as a “fan” term, and uses it to specifically refer to mirror scenes in Lost’s final season as “series-long echoes” (2004, 58). TVTropes simply defines a callback as “A relevant reference to an event taking place earlier than the timeline of the present story”. TVTropes suggests that a call-back is the “opposite” of foreshadowing, which would suggest that we can equate a callback with an analepsis in narratological terms. But even de Jong’s exhaustive narratological approach to the Iliad is quite vague in defining analepses, using only Genette’s distinction between internal and external analepses, with internal referring to plot points within the narrative, and external referring to those before the narrative’s scope (de Jong 1987, 82–85). Within this dichotomy, we would consider all television callbacks to be internal analepses.

In examining internal analepses in the Iliad, de Jong uses an example from Achilles’ killing Hektor, line 22.323, as Achilles eyes up Hektor’s armour before attacking him: “the lovely arms that (Hektor) had stripped from Patroklos when he killed him”, which, as de Jong points out, “recapitulates the very reason why Achilles is so eager to kill Hector” (de Jong 1987, 87). But de Jong does not speak of the repetition of the lines that further link Hektor’s and Patroklos’s death which I mentioned above: how do we understand that kind of reminder in relation to de Jong’s clearer example of an internal analepses? If they are both callbacks, and I think that they are, how do we distinguish between the two, both in terms of type and in terms of narrative and affective meaning?

These kinds of questions remain when we think of callbacks within a broader context of television recall strategies. Mittell’s “Previously On…Prime Time Serials and the Mechanics of Memory”, offers the following television elements which could be equated to analepses:

1. diegetic retelling – character dialogue that recalls past narrative events;
2. voice-over narration;
3. flashbacks (with additional extra-narrative recall strategies coming from title sequences, recaps) (Mittell 2010, 81–89).

In his 2015 Complex TV, Mittell likewise understands television recall strategies as "diegetic retelling", and "more subtle visual cues such as objects, setting, or shot composition can serve a similar function to activate long-term memories" (2015, 182). We can expand on this taxonomy by looking at much of the work done in analyzing television callbacks by the popular press and fans. Just reviewing a few of the articles on the callbacks from Game of Thrones’ finale (Chaney 2019; Renfro 2019; Robinson 2019; Van Boom 2019) gives us a rough list of callback types that expands on those established by Mittell (in italics here), but is also specific to Game of Thrones (2011):

1) diegetic retelling

2) voice-over narration

3) flashbacks

4) repeated emphasis on an object

5) repeated emphasis on a setting

6) repeated shot composition

7) repeated scenario (type-scene)

8) repeated shot composition

9) mirror shots/ what TVTropes would refer to as “Book Ends”, which correspond to epic theories of ring composition and encompass Lavender-Smith’s discussion of callbacks in Lost (2012, 58)

10) fulfilled predicted action

11) repeated musical cue

12) repeated costume element

13) returning characters

14) reiterated character traits

15) repeated language

This initial list begs for further research, both in refining and expanding this taxonomy, as well as asking substantive questions around this taxonomy’s significance in storytelling across genres and media platforms. Moving forward, in both epic and television studies, I hope to further examine these types of callbacks, when they are used, how frequently they are used, and how their uses create different kinds of audience affects.

Epic Episodes

One reason for continued examination of these recall strategies, or callbacks, lies in their relationship to episode structure. Considering episodes as a narrative unit provides one of the most difficult points of comparison between ancient epics and contemporary North American television, as I have stated before. But at the same time, the new norms of season-drops and tendencies towards allowing storytelling to dictate episode length, rather than episodes following pre-determined commercial constraints, also show a new “epic” flexibility in understanding narrative units. Platform matters, here, too: Baker goes so far to say that the Netflix model “sets up a viewing experience in which any television serial or episodic/serial hybrid is conceptualized as one whole, expansive, epic text” (2017, 48). Ancient epics don’t have episodes per se, although of course they would have had to have been broken up in order to have been performed or to have been watched: no one can binge 27 hours of anything, and more, there’s no way that anyone could ever perform for that long. O’Sullivan talks about both television and epic poetry’s “segmentivity”, seeing them both as “broken on purpose” (O’Sullivan 2010, 60). Still, television – even full season-drop television in its “disrupted” seriality (Buonanno 2019, 194) – still gives us episodes as a unit (VanArendonk 2019), meaning that the narrative structure still takes narrative breaks regardless of where or if the viewer might. But how do we figure out where the breaks are in epic? How do we define epic episodes?

The problem of epic “episodes” has existed for millennia: the transmission of the Iliad and the Odyssey has been, since the Middle Ages, in twenty-four “books” or “rhapsodies” of vastly varying lengths: the shortest is 461 lines (Book 5), the longest 897 lines (Book 23), or about 38-75 minutes of performance time. This difference in itself might suggest a case against understanding the books as episodes, but as already stated, contemporary North American television has started to a similar range in terms of episode lengths: The OA’s first season (2016), for instance, has episodes that range between 38 minutes and 71 minutes (Sciretta 2017); Horace and Pete’s single “season” (2016) has episodes ranging from between 30 and 67 minutes (Kozak 2016, 10).

Contemporary debates around the Iliad’s book divisions centre around two questions: how old they might be and if they correspond to performance breaks. Jensen is persuaded that they date back to the 6^th century BC and are actually a facet of the poems’ original transcription under the Peisistratid tyranny at Athens (Jensen 1999), but most scholars see them as later additions (Taplin 1992). Whether or not these book divisions are “original” connects to whether or not they represent performance breaks: some scholars say that they all do (Stanley 1993, 249–61); others think only some do (Taplin 1992); and still others think that line structure of the Iliad allow for breaks almost anywhere, but do not necessarily suggest any concrete breaks (Edwards 2004, 58). These questions around book divisions can be taken in parallel with those around analogous performance traditions, in twentieth-century Turkey and Yugoslavia, where, again, songs could last between a half hour and many hours. Scholars also look towards ancient literature for clues as to how the epics might have been performed, many leaning towards shorter “episodes” (Burgess 2004, 8n.34), like the one that the rhapsode Demodokos performs in the Odyssey (Book 8); (Ford 1997, 85): this is in part based on the ancient custom of referring to Iliadic “episodes” by their titles, like Herodotus talking about “the aristeia of Diomedes” (our Book 5); (Hdt. 2.116); (Ford 1997, 113), or Plato’s Socrates talking about “the prayers” (our Book 9; Hipp. Min. 364e-365a). But as Ford points out, Herodotus’ reference actually takes place in our Book 6, another clue that the book divisions might be a later addition (ibid.).

This impossibility of pinpointing episodes within the ancient epics made me hopeful that by thinking about television episode structures, we might have a better sense of how the Iliad and the Odyssey might best be divided for audience comprehension. Looking at a variety of contemporary North American television shows, I did beat-by-beat close readings to see how shows built recall strategies into their narratives. After a cursory study, which does not yet reflect the nuances of taxonomy that I suggest above, all the shows that I looked at, whether weekly or season-dumped, had clear patterns of repetition, where each episode generally explicitly recalled at least one narrative event from each proceeding episode (at least within a single season), while also reiterating primary plot points: for instance, this is true for both Game of Thrones (2011) and The OA (2016). Stranger Things (2016) and Twin Peaks: The Return (2017) shows similar patterns, but both have exceptions as well, in their so-called “standalone” episodes: Stranger Thing’s “The Lost Sister” and Twin Peaks’ famous “Part 8”; though even “The Lost Sister” shows a remarkable amount of narrative recapitulation (Zeller-Jacques and Kozak, forthcoming). While I have not yet looked at television shows for type-scenes, that, too, would be a fruitful avenue of research: I would guess that most shows only repeat specific type scenes once per episode, with the exception being comedy, which relishes repetition³.

The Iliad shows us something a bit different in terms of its recall strategies and how it leans into repetition. Many narrative events are never recalled, and many “episodes”, however we might divide up 40 to 70 minute chunks of the text, never refer to preceding narrative events, perhaps lending credence to the idea that it was not meant to be performed whole (Ford 1997). Even when we think of major plot points, like the argument between Achilles and Agamemnon that sets everything in motion in the first book, at one point goes over a thousand lines, or between ninety and a hundred minutes, without any mention (10.104-11.607; 1082 lines). After their argument is resolved in Book 19, we have an even larger gap without mentioning it: over twenty-five hundred lines, or nearly four hours of performance time (19.259-23.884; 2679 lines). Contrast this kind of major plot point recapitulation to Stranger Things 2’s (2016) reiteration that there’s a shadow monster over Hawkins: even the simple image of that hovering monster reappears at gaps within the show of no longer than around sixty-five minutes, and appears in every episode except “The Lost Sister”; Game of Thrones’ (2011) first season takes an average of thirty-five minutes to remind us that the White Walkers have returned; Game of Thrones’ seventh season reiterates that the army of the dead is on the march at least twice in every episode, whether through showing the army of the dead itself (episodes 1,5, 6, and 7) or through diegetic retelling (every episode). Jon Snow himself says some variation of “the army of the dead are on the march” at least eight times just in the seventh season’s nearly eight hours. Achilles retreads his fight with Agamemnon fewer times in the entire Iliad.

Moving Forward

All of this leaves us with more problems than satisfying conclusions. The Iliad’s seemingly lower-than-television rates of recall strategies might come as a surprise to Homeric scholars, and might suggest several things: first, the paratactic character of its composition, where certain narrative events might not be recalled simply due to the interest of new poets or audiences (Notopoulos 1949); second, the traditional knowledge of the Iliad’s audience, who might know more of the Iliad or the Iliad’s events than a television audience coming to a new storyworld (Scodel 1997); third, what we think of as “episodes” may indeed be longer, even much longer, chunks. This could reflect past scholarly assertions about the Iliad’s “tri-partite” structure, with performance sections coming in between six and ten hours (Heiden 1996): the extensive recapitulations of the primary conflict between Achilles and Agamemnon in Books 9 and Books 16 could then suggest that those are beginnings of new performance stretches. Just the same, my reluctance to insist on such long chunks has perhaps only strengthened after my experience performing the Iliad in shorter weekly episodes of around an hour – as I move forward in performing longer chunks of text, my view on this might change (Kozak 2016, 3). After taking all this into consideration, I might be tempted to follow Ford (Ford 1997) and argue for one-off episode performances, but there is still some sense in me that feels the narrative accumulation of the Iliad is too great to be ignored: all those battle scenes might work as one-offs, but feel much more convincing as part of a narrative expansion that gives the contraction of the epic’s final third much greater weight. And so despite my best efforts, both through comparative poetics and through performance, the Iliad remains elusive in its performance model.

When we return to television, the huge rate of recall strategies, even for season-dump shows, might surprise television scholars, especially those who see us moving into more elevated narrative strategies as North American television has moved onto cable and streaming platforms (contra (Warhol 2014; Burroughs 2019; Buonanno 2019, 190). In fact, what might be changing about North American television narrative strategies in the cable/season-dump model is not the number of recall moments or even the amount of diegetic retelling, but instead diegetic retelling integrating more smoothly into scenes that no longer need compensate for commercial breaks, and more varied types of recall strategies combining at greater rates to create a more immersive storyworld experience that may match the possibility of viewer bingeing (Zeller-Jacques and Kozak, forthcoming).

So in comparing television and ancient epic, we can still find use in seeing both as media for long-form non-literary narratives that rely heavily on narrative recall strategies and repetition, but I think the analogy between the two might have reached the limit of its usefulness, at least for me, for now. That being said, there is much to do on both fronts moving forward. As mentioned earlier, both forms need greater specificity of analysis in terms of type-scenes, and better articulated taxonomies of recall strategies more generally. Television studies could certainly benefit, too, from the kind of close reading (cf.  Hudelet 2020), digital archiving and text mining that the Homeric texts have benefitted from, especially given the enormous amount of audiovisual “text” that television is producing. For the time being, this labour largely rests in the hands of the popular press, and, even more so, of fans, whose dedication from finding “Easter eggs” and callbacks to creating supercuts of repeated phrases or shots has humbled me in my early television research and must not go unmentioned⁴. Still, to truly understand how television narrative works, and how and if it is changing as our viewing platforms and practices change, these kind of massive data sets will be essential tools for analysis. So looking forward presents a whole horizon of possibilities, of new understandings, and of greater appreciation for these stories that continue to captivate us.