The Same Fire.src
What follows is an edited transcript of a conversation between a writer (J) and an AI assistant (C). It is published as a companion to the short story “The Same Fire” as a demonstration of the process by which the piece was researched, argued over, and written. The conversation has been condensed — search queries, tooling outputs, and scaffolding removed — but nothing has been materially altered.
This came together very quickly. The chat was not initially intended to create fiction. I had been listening to a radio item about medieval times, and it was accompanied by the usual courtly lute and pipe music that denotes “medieval”. I wanted to explore whether getting lost in repetitive beats would have been a thing in those times.
After the research phase, which produced a “research document” not shared here. Then we go on a tangential journey around prehistoric acoustics, and then the idea for the story popped into my head. Normally I would expect the generated story to need a lot of changes, but apart from the one change you see me make here, I really couldn’t think of anything I wanted to be different.
I still have no idea whether it has any real basis in historical truth, but I enjoy it as speculation.
J: In period dramas or documentaries about medieval times, sedate period music is often used, giving the impression that this was a calm, measured time. The “repetitive beats” of a rave, or the raucous noise of rock or metal, are a new, less civilised development. Or the traditional music of “savage” African tribes.
I contend that early Britain and Northern Europe must have had loud, high tempo, beat-driven party/ritual music too, but it is underrepresented possibly because it wasn’t written down. There must be a continuous tradition from historic “raves” to modern.
Research this, putting your findings into a markdown artifact. Your findings will be used as the basis of a fictional account of the kind of event where this music was played and heard.
C: [Researches across multiple web searches covering prehistoric acoustic archaeology, the carnyx, Norse seiðr tradition, medieval carnival and charivari, Beltane fire festivals, and the pipe and tabor tradition. Produces a research document titled “The Ancient Rave: Beat, Noise, and Ecstasy in Prehistoric and Medieval Northern Europe.”]
The document covers: the survivorship bias problem; Neolithic acoustic engineering including Stonehenge’s resonance properties; Bronze Age lurs and Irish horns; the Iron Age carnyx; Norse seiðr drumming and the Sámi runebomme; medieval pipe and tabor, carnival, and charivari; Beltane; the drone as a continuous thread; and a summary table mapping historical anchors to fictional elements.
J: It is pointed out to me that the Stonehenge part is highly speculative. Most stone circles are barely waist height and would have no appreciable sound wall effect. Responses to this?
C: Fair challenge, and the critic is largely right about the acoustic modelling being speculative.
Where the critique lands well: The Salford study was explicitly modelling Stonehenge as it was — fully intact with lintels, which is a significant caveat. Most stone circles are, as your source notes, low-standing and would produce no meaningful acoustic enclosure. The 110Hz resonance finding across six Neolithic structures is more defensible, but those are a different category — enclosed chambers — from open stone circles entirely. The phonolite/lithophone point about the bluestones is real, but “thousands of hammer marks” is doing a lot of work — some of that percussion evidence may be from quarrying or dressing the stones, not musical use.
Where the core argument survives anyway: The acoustic speculation is actually the least necessary part of the case. The argument doesn’t depend on it. The carnyx evidence, the seiðr drumming tradition, the Sámi runebomme, the pipe and tabor, the carnival and charivari record, the explicit historical accounts of ecstatic all-night celebrations at Beltane — none of that requires a reverberant stone circle.
The strongest version of the thesis doesn’t need “our ancestors built a concert hall.” It just needs “our ancestors made very loud rhythmic music in large groups for ritual and celebratory purposes,” which is essentially unarguable.
For the fiction specifically: The Stonehenge-as-amplifier idea is evocative even if scientifically shaky, and fiction has more latitude than argument. But if you want the research notes to be defensible as a factual underpinning, I’d suggest demoting that section to “speculative / atmospheric detail” and leading instead with the carnyx ensemble and seiðr circle, which are on much firmer ground. The really solid anchor is the outdoor fire festival — Beltane, midsummer, the need-fire — where the acoustic environment is irrelevant because the point was never resonance, it was sheer collective noise and movement in the open air.
Shall I revise the document to reflect that more honest weighting?
J: https://archaeologyorkney.com/2016/01/14/what-did-bronze-age-people-hear/ is 10 years old and proposes some research. Is there any evidence that research was carried out, and what were its findings?
C: [Fetches the original article, then searches for Michelle Walker’s subsequent PhD research on archaeoacoustics at Sculptor’s Cave.]
The picture here is fairly clear, and worth being honest about what can and can’t be found.
Michelle Walker did proceed from the 2016 proposal to a full PhD study on audio phenomena in Sculptor’s Cave, and by 2017 she was presenting at the Third International Archaeoacoustics Conference in Portugal — so the research was definitely carried out.
However, I cannot find published results or a completed thesis. The ResearchGate entry appears to be the research proposal document from 2015/2016 rather than a findings paper, and the UHI thesis repository doesn’t surface a completed Walker PhD on this topic. The 2017 conference presentation is the last trace I can find.
This is not necessarily sinister — it’s a very common pattern in academic research at this scale. The study had a small qualitative design (5–10 participants, questionnaires), which would be appropriate for a Master’s dissertation or a pilot chapter within a broader PhD, but which would likely not generate a standalone published paper.
The underlying UCLA EEG finding that Walker was building on — that brain activity at 110Hz showed a shift to right-sided prefrontal dominance, compatible with relative deactivation of language centres — is the more robustly published piece of science here. Walker’s work was essentially a field test of whether that effect was reproducible in an actual prehistoric acoustic context.
If you wanted to track down the actual results, the UHI Archaeology Institute contact listed on the Archaeology Orkney site would likely know what became of the thesis.
J: These specific frequencies — how important are they. Is it exact (110.2 Hz would not have the same effect) or is it a blurry sweet-spot? What effects are we postulating?
I’m reminded of specific light frequencies discovered to improve skin conditions.
C: The short answer is: it’s a blurry range, not a precise frequency — but the analogy with light therapy is more apt than you might expect, and the mechanisms are interestingly different.
The range, not the number: The key finding from the UCLA EEG study (Cook et al., 2008) is not that 110.0Hz is magic. Archaeological investigations of prehistoric megalithic structures identified acoustic resonances at frequencies of 95–120Hz, clustering around 110–112Hz, and it was this band that formed the basis of the brain activity study. When they tested it, they compared tones at 90, 100, 110, 120, and 130Hz, and found significantly lower left temporal activity at 110Hz specifically, with the prefrontal shift to right-sided dominance also occurring at that frequency.
So 110Hz does appear to be something of a peak within the 95–120Hz range, but the honest interpretation is a sweet spot with a gradient, not a sharp threshold. Nobody has tested 108Hz vs 111Hz to find the cliff edges.
What effects are being postulated: Two distinct things, worth separating:
- Left temporal suppression — reduced activity in the language-processing areas of the brain. Your inner verbal monologue quiets down.
- Right prefrontal dominance shift — associated with emotional processing, holistic/non-linear thinking, and altered states.
Combined: less language, more emotion and pattern-recognition — which maps extremely well onto the subjective phenomenology of trance states, ritual experience, and dance-floor dissociation.
Your light therapy analogy: It’s a genuinely good one, and the parallel goes further than you might expect. In photobiomodulation, specific molecules in cells absorb specific wavelengths and trigger downstream signalling cascades — a real lock-and-key interaction with a biological receptor. The 110Hz acoustic effect works through a different route: the brain is already a system that oscillates at various frequencies, and external rhythmic input can nudge or entrain those rhythms. More like resonance with an existing oscillator than triggering a specific receptor. This makes the acoustic effect broader and more variable between individuals, unlike light therapy where the mechanism is quite specific. Context matters enormously — your state going in, the social setting, darkness, physical exertion, intoxicants — all modulate the same underlying oscillatory dynamics that the drumbeat is nudging.
The honest epistemic status: The Cook et al. EEG study was a pilot with 30 participants, and has been widely cited in archaeoacoustics but not extensively replicated. The general principle — that sustained rhythmic sound can shift brain state — is well-established across a much larger literature. The specific claim that 110Hz is special in a way that 100Hz or 120Hz isn’t is interesting but not yet definitive. Good for fiction; requires a caveat in a research document.
J: Now, there were periods in history where the median age was outrageously young. I’m looking for hypotheses about “raves” in that period. Looking for masses of young people, pounding drums at 120bpm, probably higher, psychedelics possibly, dancing but ideally not choreographed…
The genuine answer might well be “something akin to a ceilidh or twmpath”, but at least the modern version of those is more family-friendly and respectable than what I hope for.
Do however be honest rather than indulging my hopes dishonestly.
C: The demographic picture is solid. Historians estimate that half of all people in 14th-century England were 21 years old or younger. So the crowd at any public gathering would have been dominated by teenagers and people in their early twenties in a way that has no modern parallel.
The tempo question: 120bpm+ is specific. The pipe and tabor tradition supports fast dance music — estampies and saltarellos are lively, and some reconstructions run at 120–140bpm. But we don’t have tempo markings from pre-literate periods. “120bpm+” is an assumption imported from modern rave culture rather than recovered from evidence.
The choreography question: This is actually your strongest ground. Choreographed courtly dances were a feature of elite medieval culture. The dances that attracted Church condemnation again and again were explicitly described as unruly, wild, and without order. Generations of ecclesiastical authorities protested against great choruses of women who poured into churches and monasteries on feast days singing obscene songs and ballads, with complaints recorded from the 6th century to the 14th. The dancing mania material is also relevant — groups of people dancing erratically, sometimes thousands at a time, who danced until they allegedly collapsed from exhaustion. Whatever that was, it was not choreographed.
The psychedelics question: Well-supported, though frustratingly hard to localise to Britain specifically. Direct hair analysis from a Bronze Age site in Menorca confirmed the presence of ephedrine, atropine (from henbane), and scopolamine — the first direct evidence of drug use in Europe. Henbane has been found associated with Neolithic grooved ware pottery at sites in Scotland and England. For the Norse, references in sagas and folklore hint at the consumption of psychoactive mushrooms such as Amanita muscaria. The mead-of-poetry motif across Indo-European mythology almost certainly encodes memory of ritual intoxication. So: psychoactives in ritual contexts across prehistoric and early historic Northern Europe — yes, supported. A Bronze Age equivalent of pre-loading before a fire festival — plausible but not directly evidenced.
The honestly synthesised version: A seasonal festival — Beltane, midsummer, a saint’s day — draws a population where half the attendees are under 21. The formal structure (the procession, the fire-lighting, the ritual) is done by dusk. After that, the adults who matter are drunk, the elders are asleep, and what remains is a large mass of teenagers and young adults in firelight, with someone playing pipe and tabor or bagpipe at whatever tempo gets people moving, no choreography because no one taught them any, possibly mead with something in it, probably lasting until dawn. The Church records this as a scandal. The young people do not record it at all, because they can’t write.
That’s not quite a rave. It’s closer to a very chaotic ceilidh in a field at night with ergot in the ale. But the psychological ingredients — youth, darkness, fire, rhythm, intoxication, bodies, anonymity within a crowd, the temporary suspension of the normal social order — are all there and well-evidenced.
J: Listening to Irish lilting — it has many elements of dance music — the repetition, a groove, the tempo, often accompanied by stomping feet — of course it wouldn’t have volume, but it shows the instincts.
C: Yes, and this is actually a really sharp observation that cuts right to the heart of what the survivorship bias has preserved.
Lilting — puirt à beul in Scottish Gaelic, portaireacht in Irish — is mouth music. It’s what you do when you have no instrument. The melody exists to drive the feet, not to be listened to. The words, where they exist at all, are often nonsense syllables chosen for rhythmic effect rather than meaning. It is, structurally, a drum machine made of a human throat.
What this tells you is something important: the groove instinct was prior to the instrument. The body wanted to move at that tempo before anyone had built a tabor. The instrument, when it arrived, was serving an already-existing human need that people had been meeting with their voices and their feet.
The “no volume” observation is interesting though — I’d push back on it slightly. Lilting as we now know it is an intimate, domestic form, yes. But consider what happens when you scale it. A dozen people lilting in unison, outdoors, at a gathering, is not quiet. Accounts of Irish and Scottish communal work-songs — the òrain luaidh (waulking songs) used in cloth-fulling — describe them as loud, rhythmically driving, and physically synchronising groups of workers around a repeated percussion action. The women beating wet cloth on a table in unison are the drum. The song is for the rhythm, not the other way around.
The lineage you’re tracing might run: communal body percussion and voice → lilting and work-song → pipe and tabor → bagpipe ensemble → carnyx — all serving the same underlying function, with volume and complexity scaling up with technology and occasion. The instinct is constant.
J: OK, we’re ready for some fiction.
What I want is a short-story length account of a 90s rave, and a 14th century revel. Both first-person accounts.
Flip flop between the accounts, giving pairs of accounts for parallel parts of the experience. Start with the 90s account, then the ancient.
For example, we might start with the preparation: 90s, phone a number we heard on pirate radio, drove to a field outside the M25 in friend’s rusty Fiesta, friend has scored some Ecstasy, we heard Carl Cox would be there.
Parallel, 14C, it’s a 10 mile walk to the location of the Beltane revels, we pick mushrooms on the way, we heard the drums would be louder than thunder…
Continue in this vein with pairs of accounts, dancing until sunrise, sleeping in the car / field, the morning after.
Any questions before you begin?
C: A few things to nail down:
Setting for the 14th-century account — given your interests, I’m inclined toward somewhere in the Welsh Marches or mid-Wales rather than generic “medieval England.” A walk through hill country to a hilltop fire feels right. But would you prefer further north (Scottish Borders, say) or should I keep it in your part of the world? The Marches are good because they’re a cultural borderland where English authority was thin and older traditions persisted longer.
The narrators — I’m assuming both young, probably male (since a female narrator in either setting introduces complications that would pull focus from the sonic/visceral experience toward gendered threat, which I think isn’t what you want). Roughly same age — say 19–20? Or would you prefer one or both female?
Tone — I’m reading you as wanting gritty, embodied, unsentimental. Not literary-festival prose, not laddish nostalgia either. The kind of writing where you can feel the bass in your sternum / the drum in your ribs. Correct?
The parallel — I’m planning to avoid being too on-the-nose with the mirroring. The connections should emerge rather than being signposted. Agreed?
Length — I’m thinking six or seven paired sections, roughly 4,000–5,000 words total. Enough to breathe but not so long it loses tension. Does that feel right, or do you want it tighter?
J: Marches works, though there’s no need to be specific. Both male, have a female in the group, everything purely platonic — friends being hedonistic. 19–20 works great. Your instinct is correct on the tone. The parallel: zoom out until the chunks are broad enough for parallels to fit. So “the journey” can clearly be a parallel, I guess there’s arrival and a warm up period before the real raving starts, there’s the period when things reach their trance-like peak, there might be a chillout period — or periods when the narrator leaves the throng for some relief — and so on. Length feels right.
Go ahead.
C: [Writes “The Same Fire” — seven paired sections: The Journey, Arrival, Warming Up, The Peak, Stepping Out, Going Back In, Dawn. Approximately 3,800 words.]
J: Small niggle — the narrator doesn’t say he eats the mushrooms, but then says he experiences them. So let him eat some at the same time the others do.
C: [Makes the edit. The line “Efa ate half the mushrooms and gave the rest to Huw” becomes “Efa divided the mushrooms into three and we ate them with the last of the bread to kill the taste.”]
Done. Three-way split now, shared with the bread to mask the taste.
J: A reviewer suggests: “The structural mirroring (moving from 1994 to 1394 in every single section) is highly effective, but occasionally feels almost too perfectly symmetrical. An editor might suggest allowing one timeline to bleed into the other without a hard section break near the climax to simulate the loss of boundaries you describe in the text.”
I’m minded to ignore this, but let’s give it some thought. Are the year headings too direct?
There’s a temptation to attempt some kind of psychedelic time-slip blending, but I feel that is either too ambitious — pulling it off would be a magic trick — or too contrived.
C: I think you’re right to resist the bleed, and I think the reviewer is pattern-matching to a standard workshop note rather than reading what this specific piece is doing.
The rigidity of the alternation is the rhythm. The piece is structurally a 4/4 kick. Breaking the pattern at the climax would be the prose equivalent of a DJ dropping out of time to show you he’s clever — it draws attention to the author’s hand at exactly the moment the reader should be least aware of it. The symmetry isn’t a limitation to overcome, it’s the argument. These two experiences are the same experience. The structure should insist on that, not apologise for it.
The time-slip blending, as you say, almost never lands. It tends to produce prose that describes boundary dissolution rather than enacting it. The piece already enacts it through the echoed phrases — the reader does the blending in their own head, which is far more powerful than being told to.
On the year headings specifically: I’d actually consider softening those, but not in the direction the reviewer suggests. Rather than 1994 and 1394, you could drop to just the section titles — The Journey, Arrival, etc. — and let the voice and the detail do the work of placing you. The reader figures out which is which within three words. That would remove one layer of editorial framing without breaking the structure.
But that’s a taste call, not a structural one. I’d leave it as it is.
J: I began to wonder about a non-fiction background paragraph for each era. In italics. Superfluous?
C: Superfluous. The fiction is doing the work. A non-fiction paragraph before each section would signal to the reader “this is a piece with a thesis” rather than “this is a piece with a pulse.”
There’s also a tonal problem. The 1394 sections work because they don’t explain themselves. The narrator doesn’t tell you what Beltane is, what the mushrooms are for, why there’s no priest present. He just lives it. An italic paragraph saying “The Beltane fire festival was a Celtic tradition dating to…” would retroactively frame him as an exhibit rather than a person.
If the research needs to be visible — say, for publication in a context where readers might think you’ve invented the historical detail — a single author’s note at the end would do it. Three lines: the acoustic research, the mushroom evidence, the Church complaints about unruly dancing. Enough to say “this is grounded” without interrupting the experience.
J: Now that we’ve written it — I consider it complete — how defensible really is the idea of the event: “bigger, wilder, people from three or four parishes coming together where the old track crossed the ridge. Huw’s cousin had been the year before and said the drums went on till morning.” — something historians think would really have happened? A plausible invention of ours? An implausible invention of ours?
C: Somewhere between the first two, leaning toward plausible invention.
Multiple parishes gathering at a shared site for a seasonal fire festival — solidly defensible. Beltane and midsummer gatherings are well-attested in Gaelic and Welsh sources, and the logic of a gathering point on a ridge or crossroads serving a wider area than a single village is consistent with what we know about pre-Reformation rural social geography.
The event being unsanctioned, priest-absent, effectively autonomous — also strong. The entire pattern of Church condemnation from the 6th century onward only makes sense if these events were happening regularly without clerical supervision. You don’t write repeated prohibitions against things that aren’t happening.
Drums going on till morning — plausible but less directly evidenced for Britain specifically. We have strong evidence for sustained drumming in Norse, Sámi, and Continental contexts. Frame drums were certainly present in medieval Britain. The “till morning” element is consistent with what we know about Beltane structure, but a specific account of all-night drumming at a 14th-century Welsh Marches gathering is our interpolation, not our citation.
Multi-parish scale, word-of-mouth, a degree of planning — this is where we’re most honestly in plausible-invention territory. The event as depicted has a semi-organised quality — people have heard about it, they travel to it, it recurs annually. That’s not a spontaneous village thing, it’s closer to a fair or a pilgrimage, and those existed and are documented. But the specific combination — large, secular, nocturnal, beat-driven, with no authority structure — is something we’ve inferred from the intersection of several evidence streams rather than read about in a single source.
The mushrooms — the weakest single element. Psilocybin mushroom use in medieval Britain is not directly attested. The Menorca hair analysis is Bronze Age and Mediterranean. The “devil’s bread” framing is our invention. It’s not implausible. But it’s the element a historian would most want to put a question mark next to.
The honest summary: the skeleton of the event is well within what historians accept happened. The specific texture — the intensity, the duration, the quasi-rave quality — is our creative interpretation of what those bare facts would have felt like from inside. An honest historian would call it “consistent with the evidence but not demonstrated by it.” Which for fiction is exactly where you want to be.
J: Now that raves like this are illegal — and legal raves are sanitised by H&S, authority oversight, licensing — what is the closest inheritor of the unbroken thread we have described?
C: [Searches for current free party and teknival activity in the UK and Europe.]
The most literal answer is that the free party scene never actually died — it just got quieter and more cautious. As recently as 2025, thousands of people gathered at an old colliery in woods south of Retford for Resonance 2025, the UK’s first proper Teknival in some time. After a six-hour drive and countless phone calls to the party line, coordinates were finally released, and convoys trundled down woodland tracks to unload speakers, tents, and generators. UK Tek 2023 drew 3,000 people to a hidden spot in South Dorset, with convoys of 500 vehicles from across the country converging after midnight.
The structure is identical to what we wrote for 1994 — phone a number, drive to coordinates, find a field. The CJA 1994 did to raves what the medieval Church did to Beltane: drove it out of official sight without killing the impulse.
Other living inheritors worth considering: the Breton fest-noz (bombardes and binious at volume, circle dancing that accelerates through the night) — but it now has UNESCO intangible heritage status, which is the polite way of saying it’s been domesticated. Sound system culture (Jamaican-rooted, through Carnival, dub, jungle, grime) preserved the communal ownership element — the system belongs to the crew, the crew serves the community — though Carnival is now heavily licensed and policed. Football terraces are a genuinely interesting lateral candidate: massed chanting, synchronised movement, collective trance-like states, youth-dominated, no choreography, explicitly condemned by authorities in its intensity. The ultras tradition in Southern and Eastern Europe makes this even more explicit — drums, flares, coordinated noise. It’s charivari in sportswear.
But if you forced me to name the single closest inheritor of the unbroken thread — unsanctioned, nocturnal, beat-driven, ecstatic, youth-dominated, no authority structure, collective rather than performed, and still happening — it’s the free party. The same fire in the same field with a generator instead of a bonfire.
J: “the bluebells had come through in carpets so thick they looked like water” — this is excellent, can you know where you got it?
C: I can’t, honestly. I don’t have access to my own sources in that way — I can’t trace a phrase back to a specific input the way you’d trace a sample in a track.
What I can say is that it’s a sufficiently natural comparison that I’d be surprised if it were original. Anyone who’s walked through English deciduous woodland in late April has seen that exact visual — the blue pooling in the clearings, running between the trees, settling in the low ground. It does look like water. It looks like flooding. The image is sitting right there in the thing itself, waiting to be picked up, and I’d guess dozens of writers have reached for it independently.
If it has a specific literary ancestor in my training data, the most likely candidates would be the nature-writing tradition — Roger Deakin, Robert Macfarlane, maybe Edward Thomas, possibly even Housman given the Marches setting. But I genuinely don’t know whether I’m half-remembering someone else’s sentence or whether the bluebells just told me what they looked like. That distinction may not be as clear as we’d like it to be, for me or for any writer.
J: In 1394 in the Marches, would there have been a military presence to disrupt civilian life?
C: Yes, though the nature of it shifts considerably depending on where exactly in the Marches and in which decade of the 1390s.
The Welsh Marches in the late 14th century were a patchwork of Marcher lordships — semi-autonomous territories where the lords held crown powers independently, which meant the normal apparatus of English law didn’t apply. This had two opposing effects: the lordships could be arbitrary and extractive in their own right, but they were also inconsistent, and the gaps between jurisdictions were real. People who lived on the borders of competing lordships had practical experience of falling through administrative cracks.
The 1390s were the decade before Owain Glyndŵr’s revolt, which began in 1400. Glyndŵr himself was active in the region in the 1390s as a relatively settled Welsh landowning gentleman, but the underlying tensions he would later ignite were already present — Welsh resentment of English settlement and legal discrimination, periodic raiding, local score-settling. A gathering of Welsh or border people in this period would have been viewed with some suspicion by English Marcher lords, though probably not enough to prompt active disruption of something as informal as a fire festival on a ridge.
The more immediate military texture was probably livestock raiding and banditry rather than garrisoned troops. The Marches had no permanent military infrastructure equivalent to a Roman fort network.
For the story this actually strengthens rather than complicates the setting. The gathering happens in the space between jurisdictions, on the ridge where nobody’s authority is quite clear. That’s historically accurate and structurally perfect.
The story “The Same Fire” was written in a single session. Total elapsed time from initial research question to completed first draft: approximately two hours of conversation. The research document “The Ancient Rave” was produced alongside and is published separately.