Chapter 1 Reading envisioned in the fourteenth and fifteenth centuries

Daniel Sawyer

A witch steals from her neighbours by sending a magic, self-moving bag to suck milk from their cows at night. Eventually, the community brings her before the local bishop. He has her demonstrate her bag-animating magic for him and orders an assistant to write down the words and actions she uses in her spell. At this point, reading, of a sort, enters the story:

The bysshop began the charme to rede,

And as she dede, he dede yndede.

He seyde and dede everydeyl

Ryght as she dede, he dede al weyl.

The sloppe lay stylle as hyt ded wore:

For hym ne ros hyt never the more.

‘Why’, seyde he, ‘wyle hyt nat ryse?

And Y have do the same wyse

And seyd the wrdys, lesse ne mo,

And for my seyying wyle hyt nat go?’

‘Nay’, she seyde, ‘why shuld hyt so?

Ye beleve nought as Y do.’¹

(The bishop began to read the charm, and did just as she did. He said and did everything exactly as she had. The bag lay still, as if it was dead: it never rose for him. ‘Why’, he said, ‘will it not rise? For I’ve acted in the same way, and said the words, no more and no less – will it not travel at my word?’ ‘No’, she said, ‘why should it? You don’t believe as I do.’)

The bishop tells the witch not to do it again but lets her go without punishment.

I take this incident from Robert Mannyng’s Handlyng Synne, a lively early fourteenth-century moral manual. For the history of imagined reading and imagined books, two facets of the passage hold particular interest: the public, aural, record-keeping quality of the imagined reading, and the practical, salvific, non-fictional thrust of Handlyng Synne itself. The poem imagines the bishop reading aloud as part of the recording of process. In his reading of the spell he attempts a speech act, for he tries to use language to make something happen. Thus the moral of the story, as Mannyng interprets it: beyond the right form of words, belief matters; if this holds true even for magic bags, it must be true for Christian practice. Handlyng Synne sits beyond fiction’s borders. Whatever Mannyng’s audience thought of the likelihood of magical milk-stealing bags, they probably understood the story as truthful in its salvific purpose, in saying something about belief. Moreover, the poem, even in a story about belief, stays unconcerned about the inner lives of the bishop and the witch: they are not characters. Not, of course, that premodern writing as a whole lacked fictionality or interiority. Criticism can trace interiority in any period, and scholars have argued for developments in fictionality in and around the twelfth century.² I only observe that plenty of imagined reading was imagined in works that, however clever and crafty in other ways – and Handlyng Synne is a clever poem – did not strive for fiction. A writer could imagine reading outside fiction, and fiction itself demands careful handling as a category.

Distance in time lends earlier literature a particular power to test categories and concepts. Descriptions of imagined reading in Middle English and Older Scots usefully complicate our ideas about fiction, remind us that reading has long meant more than just one person looking silently upon a book, and show how a book might take a variety of forms and have porous borders. Here, I shall explore one literary-critical touchstone for imagined reading in the period, and then trace what happened when a series of poets turned the idea of reading followed by visionary experience into a tradition. Once it became a recognisable topos, imagined reading leading into a vision allowed writers to inflect and position their work through the precise connotations of their language for the physical book and the act of reading; recent and older work exploring this area of vocabulary in the period lets us explore that bookish language.³ The developing idea of private reading appears in literary accounts. But so do counter-teleological examples of reading imagined in non-private, inhabitable, communal ways. These works show why and how the history of imagined reading might think carefully about the concepts of fiction and of the book, and might avoid stories of assumed progress.

One might call imagined reading in Scots and English at this time unimportant. Latin was the normative written language, and intellectual works in Latin had a carefully worked out set of approaches to reading, recoverable today in manuscript evidence and in surviving overt discussions. Larger-scale histories of reading have mapped out these materials, and have identified major changes in the twelfth and thirteenth centuries: an increasing sophistication of on-page aids and a growth in wide-ranging scholastic reading.⁴ Furthermore, besides Latin, the island of Great Britain sat well within French’s cultural sphere of influence: England had been conquered by a Francophone Norman dynasty, who had gone on to conquer Wales, while Scotland frequently found itself allied with France against that same dynasty. French consequently persisted as a prestige vernacular. Indeed, it is if anything in this period in particular, roughly from the eleventh century to the fifteenth, that we find Scots and English at their most decentralised, peripheral and non-standardised.⁵ One might, then, find good reasons to focus these questions on Latin or French materials.

However, the very peripheral status of English and Scots itself demands attention. It was during this period that writings in these tongues were at their most uncertain and experimental. Genre, voice, fictionality and value were ‘in play’ in ways they never quite would be again. Writing was both transnational and regional, but rather less national than it has been since. The period therefore offers a vital standpoint from which to encounter writings from later times: less national, less standard, not dominant and not an assumed default. By studying Scots and English in this period we might slide out from underneath some assumptions about fiction, reading and books.

It is in one of the Middle English works long lauded for its quasi-novelistic attention to inner life, Geoffrey Chaucer’s Troilus and Criseyde, that the period’s most often-noted moment of imagined reading occurs.⁶ Having come to Criseyde’s house, her uncle Pandarus:

… fond two othere ladys sete and she,

Withinne a paved parlour, and they thre

Herden a mayden reden hem the geste

Of the siege of Thebes, while hem leste.⁷

(Found her sat with two other ladies within a paved parlour; the three of them were hearing a maiden read them the tale of the siege of Thebes, for as long as they liked.)

When Pandarus asks about Criseyde’s reading material, she reports that:

This romaunce is of Thebes that we rede;

And we han herd how that kyng Layus deyde

Thorugh Edippus his sone, and al that dede;

And here we stynten at thise lettres rede –

‘How the bishop, as the book kan telle,

Amphiorax, fil thorugh the ground to helle.’

(We’re reading a romance about Thebes, and we have heard how King Laius died through his son Oedipus, and all that story; and we’ve stopped here at these red letters: ‘How Bishop Amphiaraos fell through the ground to hell, as the record can tell.’ II.100–5)

Criseyde and her companions are reading together, as we know people commonly did in the fourteenth and fifteenth centuries, and would do in later times too.⁸ In modern terms, Criseyde listens and it is the ‘mayden’, junior to the others present, who reads; to Criseyde, though, they are all reading, and the poem imagines her actively engaged with the material book too, identifying the red ink topical rubric which lets readers track their position, and which the maiden will presumably use to find their place again. I suspect the last two lines quoted suggest that Criseyde reads the rubric out, and so I have imposed quotation marks on them in reproducing them here.

Scholarship’s conception of reading must be open to groups, who might relate only intermittently with the physical book. The word read and its cognates do not, in their origins, assume contact with writing. They originate in a verb for interpretation and counsel, senses which they kept through Old and Middle English. Etymology is not destiny, but these meanings do persist today: when critics speak of ‘a reading’ of a work, or when a pundit gives their ‘read on’ geopolitics, the word’s oldest senses sit foremost.

Criseyde’s household reading also leads her and Pandarus into questions of gendered bookishness, and perhaps gendered linguistic competence. Both Criseyde’s description of a ‘romaunce’ that ‘is of Thebes’ and the French form of Amphiorax rather than Amphiaraus hint that Criseyde’s group are reading the Roman de Thèbes, a French version of the tale. Pandarus, meanwhile, remarks that of this story ‘ben ther maked bookes twelve’ (‘twelve books have been written’, II.108), suggesting that he knows Statius’ twelve-book Latin Thebaid. Chaucer’s characters briefly sketch an awareness of the reading boundaries between Latinity and Francophony, and the hazy borders of history and romance. Later in the poem, Criseyde’s reflections on divine power suggest wider and more intellectual reading.⁹

Such moments, though, work as lone allusions and descriptions. What happened when imagined reading became a topos, a tradition? In fourteenth- and fifteenth-century Great Britain, dream-vision poems developed a tradition of opening with imagined reading. As each poet reformulates the idea of reading before vision, this line of works offers helpful case studies. Criticism has said much about literary allusion in these passages, but less about the specific wording of the descriptions of reading and of books.

Chaucer seems, in the surviving record, to have invented the idea of describing the reading of a particular book that then affects the ensuing dream.¹⁰ Being Chaucerian, the relevant passages in the Book of the Duchess and the Parliament of Fowls have been well picked over for their allusions and the roles they play in their respective poems.¹¹ The Book of the Duchess begins with a speaking, narrating figure who cannot sleep:

Upon my bed I sat upright

And bad oon reche me a book,

A romaunce, and he it me tok,

To rede and drive the night away;

For me thought it better play

Then playe either at ches or tables.

(I sat upright in my bed and asked someone to hand me a book, a romance, to read and drive the night away, because it seemed to me a better amusement than playing either chess or backgammon, and he brought it to me, 46–51.)¹²

Reading for Chaucer’s sleepless speaker reveals a social position: though – judging by his uncomprehending conversation with the knight encountered later in the poem – not presented as courtly, he has servants enough to be non-specific about them (‘oon’), and to have his reading matter fetched. The book request might have a haphazard tone: the reliance on someone else, the indefinite article, and especially the baggy term romaunce hint at this. ‘A romaunce’ only shakily describes what the speaker in fact receives, which turns out, as the reader relates its plot (62–230), to be Ovid’s Metamorphoses. The sense of Middle English romaunce is a hot potato of a topic, but the word typically means a knightly adventure story, more rarely simply a story or poem, or, more rarely still, a work in French.¹³ Metamorphoses might be filed under the second of these senses, or under the third in French translation, but it fits the noun a little awkwardly, and Chaucer might have hoped an audience would be piqued and amused by realising that the speaker’s haphazardly fetched ‘romaunce’ is Ovid. That effect would, at least, fit with the way in which the speaker’s limited perceptions drive comedy and pathos later in the poem.

When sleep does come for the poem’s dreamer, it brings him and the book together: ‘Such a lust anoon me took / To slepe that ryght upon my book / Y fil aslepe’ (‘at once such a wish to sleep took hold of me that I fell asleep right there, upon my book’, 273–5). This is personal, embodied reading. The wording could have a faint metaphorical edge, for to sleep ‘upon’ something in late Middle English could, somewhat rarely and mostly in the fifteenth century, mean to sleep with an idea in one’s head (MED, 1(d)). Chaucer’s wording might therefore invite that sense too, letting an audience hear both senses: physically sleeping with the book, and being influenced by its contents.¹⁴ The poem’s end reprises the same idea: the tolling of midnight by a bell inside the dream wakes the dreamer, who finds himself lying in bed, ‘And the book that I hadde red, / … I fond hyt in myn hond’ (‘and I found the book that I had been reading in my hand’, 1326, 1329). Again Chaucer’s wording emphasises the bodily connection here, but from this and from the earlier mention of the book an audience might also reasonably imagine a portable book. The book can be read in bed, and it can be held in the hand. Aspects of this sound very normal, and we need not downplay ways in which Chaucer’s description dovetails with present-day reading in bed. Nevertheless, it does remain part of the studied, mannered dream-vision genre: these passages do not require readers to imagine a plausible, consistent fiction in which Chaucer’s book must be one definable size, but they do craft a casual, personal model of reading.

The opening of Chaucer’s Parliament of Fowls strikes some similar notes:

Nat yoore

Agon it happede me for to beholde

Upon a bok, was write with lettres olde,

And thereupon, a certeyn thing to lerne,

The longe day ful fast I redde and yerne.

For out of olde feldes, as men seyth,

Cometh al this newe corn from yer to yere,

And out of olde bokes, in good feyth,

Cometh al this newe science that men lere.

But now to purpos as of this matere:

To rede forth hit gan me so delite

That al that day me thoughte but a lyte.

(Not long ago, I happened to see a book that was copied in old handwriting, and I read in it all the long day, very fast and eagerly, in order to learn a certain thing. For, as men say, new corn comes from old fields year on year, and, truly, all the new knowledge that men learn comes out of old books. But now to the point of this discussion: reading on delighted me so much that the whole day seemed to me just a little time, 17–28)

Chaucer, like his contemporaries, often invoked old books, and the link between ‘olde bokes’ and ‘newe science’ elaborates on a routine reflex.¹⁵ Chaucer did not often, though, ask readers to imagine the age of handwriting, and his remark that the book ‘was write with lettres olde’ makes readers imagine a material quality of the book described. The scripts taken as models in book production changed over time in observable ways – thus the existence of palaeography as a field – and remarks such as this one in verse, and many others in surviving practical writing, show that premodern readers themselves knew these changes.¹⁶ Chaucer writes nothing more precise that would guide the imagination towards a book copied in a hand modelled on (say) a much earlier and radically distinct script such as English Caroline minuscule, or merely a book copied in a hand modelled on an earlier form of a more familiar script. But his remark on the ‘lettres’ sets his audience up to envisage books in which every line makes clear the age of the leaves.

Besides a hint of everyday palaeography, this passage also indicates a particular kind of reading. It is deeply involved:

To rede forth hit gan me so delite

That al that day me thoughte but a lyte. (27–8.)

To underscore the wholly engrossing, time-swallowing quality of the reading, ‘day’ holds the same second-beat position in both of the stanza-terminal lines:

The lónge dáy ful fást I rédde and yérne

That ál that dáy me thóughte bút a lýte.

The reader ceases only when nightfall cuts off his light source (85–7). The poem describes an episode of what the history of reading in a later period might call intensive reading. Having been reading Somnium Scipionis, itself a dream-vision text, the reader falls asleep and dreams a complex, finely worked vision of artifice, interpretation and debate. The poem’s close takes a different tack to the physicality and specificity of the waking upon the book in the Book of the Duchess. The dreamer moves precisely not to the book that he was reading: ‘I wok, and othere bokes tok me to’ (‘I woke, and turned to other books’, 695). The Parliament ends in reading unlike that found in the Book of the Duchess, reading that is hunting, multi-codical, and perhaps less physically engaged with the material page. One poet could make different uses of the sequence of reading and then dreaming. So much for Chaucer’s initiation of the tradition of books prompting visionary reading. Where did that tradition go?

The next major work to use the book-before-vision topos is The Kingis Quair. Internal biographical details and the title – an external imposition – strongly suggest that James I of Scotland (1397–1437) wrote this poem, perhaps in the 1420s.¹⁷ The poem begins with its dreamer around midnight:

Quhen I lay in bed allone waking,

New partit out of slepe a lyte tofore,

Fell me to mynd of many diverse thing –

Of this and that – can I noght say quharfore,

Bot slepe for craft in Erth myght I no more.

For quhich as tho coude I no better wyle

Bot toke a boke to rede apon a quhile.

Of quhich the name is clepit properly

Boece (efter him that was the compiloure)

Schewing the counsele of Philosophye …

(When I lay awake and alone in bed, having just woken from sleep a little before, I thought of many different things, of this and that; I can’t say why, but I couldn’t get to sleep by any earthly method. I knew no better trick than to take a book to read for a while – a book properly titled Boethius, after its compiler, that explains the advice of Philosophy [i.e. Boethius’ De consolatione philosophiae].)¹⁸

The dreamer reads not just at night, but after sleeping. This passage’s account of waking from an earlier sleep might, perhaps, describe the first part of a premodern biphasic sleep pattern.¹⁹ The poem presents reading Boethius as a sustained, time-bound experience. Like the dreamer in the Book of the Duchess, the protagonist of the Quair – let us call him ‘James’ – places the book at his head as he returns to trying to sleep, but he also feels the physical toll paid by late-night readers, so that his reading is if anything even more embodied:

The longe night beholding (as I saide),

My eyen gan to smert for studying.

My buke I schet and at my hede it laide

And doun I lay.

(Having gazed through the long night (as I said), my eyes smarted from studying. I shut my book and laid it next to my head, and down I lay, 50–3.)

The poem crafts a strikingly close physical identification of reader and book through the return of the same verb, first as a transitive with the book as object (‘laide’), then intransitively (‘doun I lay’). The pattern of following events differs from those in Chaucer’s dream visions; though older criticism sometimes took Scottish poems such as the Quair as Northern reflexes of Chaucerianism, they stand rooted in a much wider range of materials than just Chaucer.²⁰ James lies awake, thinking, rises at Matins, has an eventful day, and only the following night begins to dream a significant dream related to his reading (510–11).

Reading returns, however, at the poem’s end: James, still distressed, rises from his dream and goes to his window, where a turtle dove brings him a comforting letter written in gold on a branch (1233–53). The turtle dove sequence happens outside and after the dream but is nevertheless knowingly fanciful. James’s imagined response deserves attention:

Ane hundreth tymes or I forther went

I have it red with hertfull glaidnese.

And, half with hope and half with dred it hent,

And at my beddis hed with gud entent

I have it faire pynnit up.

(Before I went any further, I read [the message] a hundred times with heartfelt gladness. I took it and, half in hope and half in fear, with good will, I have pinned it up neatly at the head of my bed, 1255–9.)

Though present-day readers probably outdo them thanks to instant messaging and email, fourteenth- and fifteenth-century readers lived in a world of non-codical reading too, sending notes to each other on parchment scraps, drafting in wax tablets, displaying texts in rooms and above doorways, peering at the graffiti in churches and so on.²¹ The Quair imagines a courtly, aestheticised offshoot of the same set of practices, inviting present-day historians of reading to keep in mind text and reading beyond the codex.

Further challenges that expand the idea of the book crop up in another Scottish poem, Robert Henryson’s Testament of Cresseid. Little is known of Henryson’s life, and the poem cannot be assigned a date much more precise than the later fifteenth century, but it draws on the same tradition. Henryson has his aged speaker reading Troilus and Criseyde:

To cut the winter nicht and mak it short

I tuik ane quair – and left all uther sport –

Writtin be worthie Chaucer glorius

Of fair Creisseid and worthie Troylus.

(To cut and shorten the winter night, I set aside all other amusements and took up a quire written by worthy and glorious Chaucer about beautiful Criseyde and worthy Troilus.)²²

The Testament offers another figure seeking to ‘cut’ the night, as in earlier versions of the book-and-vision topos, but in this case one who selects his reading matter himself, with intention. He picks up a ‘quire’, and this word, already encountered in the epitextual, retrospectively imposed title of The Kingis Quair, now demands attention. Modern codicology uses ‘quire’ to mean several folded sheets (bifolia) tucked one inside the other. Such groupings are the building block of the manuscript book and the atom of codicology. In Middle English and Older Scots, though, the term had a wider and looser range of meanings. It might indicate a short work (as in The Kingis Quair), or an unbound book.²³ Wills, booklists and other writings about books from the period often mention books ‘in quires’ or simply as ‘a quire’, and such mentions seem to mean books living without hard bindings, either ‘limp bound’ in flexible parchment, or loosely gathered as bundles of quires.²⁴ In fact, most manuscripts that survive do so in much later bindings, and scholars can only rarely prove that a given manuscript had a binding within the period; though twenty-first-century pop culture tends to imagine a weighty tome with a thick wooden hard-board binding when given the phrase ‘medieval manuscript’, it is possible – plausible – that loosely bound or unbound books were the more common norm, as paperbacks are today. In choosing the word, then, Henryson gestures towards fluid, mobile reading in informally gathered bundles of leaves, not the meditative, devotional contemplation of one great work/book. While criticism has typically taken these lines as a description of reading Troilus and Criseyde, ‘ane quair’ might even mean just one physical group of leaves from a copy of the poem. For the narration of the Testament goes on to specify that:

thair I fand efter that Diomeid

Ressavit had that lady bricht of hew,

How Troilus neir out of wit abraid

(there I found how, after Diomede had taken in that bright-complexioned lady [Cresseid], Troilus leapt almost out of his wits, 43–5).

These lines refer to Book V of Chaucer’s poem, and so Henryson might well ask readers here to imagine one specific physical part. He will go on to inflect Chaucer’s story of Troilus and Criseyde by recounting his own version of subsequent happenings. By invoking the physically looser world of un- and lightly bound books, he can prepare readers for a mythologically looser world of stories open to change in the hands of different poets. Henryson’s precise wording for imagined reading in an imaginary book makes up part of his broader artifice.

Henryson continues writing of Troilus:

Of his distres me neidis nocht reheirs,

For worthie Chauceir in the samin buik,

In gudelie termis and in joly veirs,

Compylit hes his cairis, quha will luik.

To brek my sleip ane uther quair I tuik,

In quhilk I fand the fatall destenie

Of fair Cresseid, that endit wretchitlie.

(I need not recite again his distress, for in the same book worthy Chaucer has given an account of his [i.e. Troilus’] troubles, in noble words and gallant poetry, for whoever will examine it. To break my sleep, I took another quire, in which I found the fateful destiny of beautiful Cresseid, who ended wretchedly, 57–63.)

The ‘uther quair’ gives Henryson a pretext to extend on Chaucer’s work. The speaker’s ability to take up ‘ane uther quair’ with ease implies – again, without needing to posit a consistent, realist setting – a book collection, a personal library. More importantly, however, this stanza brings the topos of imagined reading before a dream vision into explicit contact with an early sense of imagined reading in fiction: the first letters of the lines read, acrostically, ‘O FICTIO’. In a study of Henryson’s intertextual play, this acrostic would send criticism off to consider his use of allusion and his claims to poetic authority and creativity. For a study of fictional reading, though, ‘O FICTIO’ asks us to consider just what fiction is, and was. Fifteenth-century fictio does not, of course, indicate quite the same thing as present-day fiction: fictio primarily meant (neutrally) shaping or fashioning, or (a little negatively) feigning and fable.²⁵ The first known written evidence for its use in Scots and English crops up in the 1490s, soon after Henryson’s writing, in the feigning sense.²⁶ If Chaucer has been fashioning or fabulating, so can Henryson: ‘Quha wait gif all that Chauceir wrait was trew?’ (‘Who knows whether all that Chaucer wrote was true?’) he asks as the next stanza begins (64). While fictio at this time was not modern fiction, Henryson’s witty deployment of it here does show a willingness to weave into readers’ experience an awareness of his own work’s construction and its place in literary tradition – and, interestingly, to weave it in a way open only to reading by eye: a listening audience would have trouble grasping the acrostic. The Testament uses the topos of reading before dreaming to loosen readers’ senses of the boundaries of books and literary works, and to show off its own processes of artful rewriting.

Reading, the book and fiction were all complex, rich ideas in fourteenth- and fifteenth-century writing. Yet the evidence presented so far might risk giving a teleological portrait, in which dream visions increasingly dabble in self-referential commentary as time goes on, moving out of religiosity into playful fiction and pointing towards the novel. I therefore close by travelling back earlier in the fifteenth century, to another example of the book-and-vision idea that offers at least as much mise-en-abyme as any of these other examples, yet turns the topos to ends not fictional but devotional.

The poet John Lydgate enjoyed a long and prolific fifteenth-century career, and positioned himself as part of a nascent Chaucerian tradition; he would later become the namesake of George Eliot’s Tertius Lydgate.²⁷ Lydgate’s ‘Fifteen Joys and Sorrows of Mary’ inherits the concept of the visionary reading opening, but uses it to pursue an unimpeachably devout meditation rather than a fictional purpose.²⁸ Lydgate draws on the same idea of serendipitous reading becoming vision. He has the poem’s visionary reader-speaker reading between midnight and morning to take their mind off the world’s troubles. They report that they ‘Vnclosyd a book’ (‘Opened a book’, 5), and the verb ‘Vnclosyd’ suggests, in its effortful litotes and its suggestion of the possibility of firm closing, a more substantially bound manuscript with hard wooden boards and one or more clasps to fasten it shut. The reading is, though, haphazard, or at least providential, rather than precise and directed:

Of fortune turnyng the book, I fond

A meditacioun which first cam to myn hond

(Leafing through the book as fortune led me, I found a meditative work that came to my hand first, 6–7).

At the start of the meditative work, the reader finds a Pietà image of Mary mourning Jesus. Then, as the reading continues, the reader begins a series of perceptive experiences that might or might not be visions:

Vpon the said meditacioun,

Of aventure, so as I took heed,

By diligent and cleer inspeccioun,

I sauh rubrisshis, departyd blak and reed,

Of ech chapitle a paraf in the heed.

(By chance, as I attended with careful and keen scrutiny, I saw, upon this devotional work, rubrics distinguished in black and red, and a paraph at the head of every section, 16–19.)

The poem merges staged reading attention and the reading apparatus expressed on the page through an elegant punning rime equivoque on ‘heed’ (16, 19), and through the subdued pun on reading and the colour red (18). Lydgate probably describes a table of contents here, and at this time tables of contents were not mundane, expected features. In a manuscript culture, when different scribal hands guaranteed that no two copies of the same work would agree in their page or leaf numbering, such tables had to be custom-made and keyed either to the specific manuscripts in which they lived, or (as here) to textual divisions: each ‘chapitle’, chapter.²⁹ This potential unfamiliarity explains why Lydgate’s description goes into some detail about the apparatus’s workings (‘Of ech chapitle a paraf in the heed’). Lydgate’s account superimposes the table on (‘Vpon’) the image of Mary, and Martha Rust has suggested in a brief discussion of this passage that he imagines the reader visually imposing a graphically highlighted apparatus onto the text from their memory.³⁰ Certainly Lydgate does give the ‘rubrisshis’ the same serendipitous arrival as the image. The picture of Mary comes ‘Of fortune … to myn hond’, and he notices the table ‘Of aventure’, by chance, the haphazardness perhaps lightly pointed up by variation into one of Lydgate’s distinctive ‘broken-backed’ lines, with its second and third beats abutting at his typical metrical caesura (‘Of áventúre, | só as Í took héed’, x/x/|/x/x/).³¹ For this poem’s opening, no neat line divides the practical route-finding table of contents from the devotional ‘meditacioun’ image in looking and reading. Readers encounter both at once, superimposed in imagined sight.

The text adds further layers of vision. In the next stanza, the visionary reader sees someone else depicted in the book, himself praying to Mary.

and to that hevenlie queene

I sauh oon kneele devoutly on his knees;

A Pater-noster and ten tyme Avees

In ordre he sayde at th’ende of ech ballade,

Cessyd nat, tyl he an eende made.

Folwyng the ordre, as the picture stood,

By and by in that hooly place,

To beholde it did myn herte good;

But of entent, leiseer cauht and space,

Took a penne, and wroot in my maneere

The said balladys, as they stondyn heere.

(And I saw someone kneel devoutly on his knees to that holy queen; at the end of each stanza he said a Paternoster and ten Aves, in order, without pausing until he had finished. It did my heart good to behold it, following the order, as the picture stood, by and by in that holy place; but, intentionally, having found opportunity and time, I took a pen and wrote in my style the said stanzas, as they stand here, 25–35.)

Claiming to (re)write the experience of vision in Lydgate’s own ‘maneere’, the poem becomes about devotion but also becomes a transmission of devotion, from the devotee in the imagined book to Lydgate and then on to the poem’s readers. This sequence offers as densely and playfully nested a set of envisioned readings as one could ask for, with the most material detail of any of my examples in its discussion of the book itself. It offers these things, though, for a different purpose, presenting itself as something readers can imitate and repeat to further their devotional relationship with Mary.

It is tempting to place these envisionings of books and reading into a historical sequence of growing sophistication and development. Past appreciations have tended to see Chaucer as an urbane, fictionalising writer, whose sophistication in describing envisioned reading flowed on to, and perhaps advanced among, erudite fifteenth-century Scots poets.³² History prefers linear process, and since the Industrial Revolution it has tended to transpose the paradigm of technical progress onto the arts. Granted, later poets do revise Chaucer’s vision, and indeed in Chaucer and elsewhere there appear flickers of reading practices very like our own, such as reading in bed. Neither the revisions nor the glimpses, however, constitute a march towards present-day reading as a telos. Rather, these poets’ envisioned readings lead in different directions from their Chaucerian origin-point. James I weaves reading more intimately into a rhythm of sleep and wakefulness, and closes his poem with the reading of notes and scraps of written material, not codices. Henryson foregrounds the role of feigning in his own activity and in his literary inheritance, and writes of the world of flexible quires, far looser in organisation and agglomeration than the unitary, fully bound early manuscript we experience or imagine today. Lydgate multiplies the layers of envisioned reading, and extends the topos of the envisioned book devotionally, with different aims and little impulse towards fiction. These poems show that diachronic changes in reading have many goals, which may pull against each other; that fiction has its own history; and that the idea of the book can fruitfully shift and multiply beyond the sealed and bound codex.

Though the recurring idea of reading and then dreaming makes up only a corner of the history of imagined reading, it might cast some light on reading’s near future. The un-teleological history laid out above should remind us that reading has a future, in the first place: reading has not reached a static, completed state, and will change around and underneath those who study its history. Reading’s future will mix newly emphasised options from among the reading approaches available since the first writings together with technological affordances which are genuinely new. The high-speed cross-referencing and the flitting quality found in present-day digital reading echo some habits of readers before the rise of the novel. The widespread present-day use of video to discuss books, especially among younger readers, effectively creates a certain kind of group reading, albeit an asynchronous kind.

However, ‘a medieval manuscript is not an iPad’, and the reading landscape today contains much that is new or at least newly proliferating.³³ Imagined descriptions and images of reading partake in a newly widespread fetishisation of the physical book, for instance, so that the material passion once restricted to eccentrics such as the fourteenth-century bibliophile Richard de Bury now pulses through broadly shared trends such as ‘dark academia’.³⁴ A strong, and in my view related, current of nostalgia can be found in the manner in which poetry is presented when circulated on Instagram.³⁵ These trends draw on a knowingly inaccurate imagination of past reading. While they involve images of physical codices, they spread on social media platforms which differ profoundly from codices, and for that matter from e-books. E-books, indeed, appear to have reached an equilibrium of only limited success.³⁶ They might have created something of a distraction from the advent of the ever-changing online feed, which has caused much greater changes in reading experiences real and imagined.

A great deal of both reading and reading about reading now takes place on endlessly updated feeds, in what has been tentatively called, adapting a term from film studies, ‘binge-scrolling’.³⁷ Feeds also channel a great deal of present-day discussion of reading, as Sarah Jerasa has observed in the teenaged readers she finds forming communities, and probably canons, on ‘BookTok’, book-focused parts of TikTok.³⁸ I do not think that Instagram and TikTok will necessarily persist for decades: online platforms come and go. But I do expect feeds like them to persist, with concomitant habits. Reading and the discussion of reading in the feed are more aural and visual – snatches of poetry come imposed on images, folk-literary-critical opinions arrive as video clips – and a feed makes reading a paratactic succession of experiences, with only a vague sense of algorithmically imposed semi-order. That is, the feed is rather like a dream. In the coming years, addressing different aspects this time around, dreams might once again help us talk about how we read and how we imagine reading.