Where the Celts Met Byzantium: Galatia, Manzikert and the Long Road East

A question that will not leave you alone

There is a moment, if you read around this music long enough, when you come across the sentence that rearranges everything. Mine was a reference to the Celts in Asia Minor in the eleventh century.

I had spent years thinking of the Celtic world as a western edge — Ireland, Scotland, Wales, Brittany, Cornwall, Galicia, the Isle of Man, a fringe pressed against the Atlantic. And here was a reference placing Celtic-speaking people in the middle of Anatolia, in the orbit of Constantinople, at the exact century when the Byzantine Empire was beginning its long decline and the Normans were landing in England.

It was not a curiosity. It reversed the geometry. If the Celts are not a western fringe but the western end of something that stretched across a continent, then the music does not have an edge; it has a road. That reversal is the foundation of everything I have written since, and The Ancestry is essentially a record about it.

Galatia: how Celts ended up in the middle of Anatolia

The history is not obscure; it is simply not taught alongside the harp. In 279 BCE a large Celtic force moved south-east through the Balkans, sacked or attempted to sack Delphi, and split. One part of it crossed into Asia Minor in 278–277 BCE, invited over by a Bithynian king who wanted mercenaries and got neighbours.

Three tribes — the Tolistobogii, the Tectosages and the Trocmii — settled on the Anatolian plateau around what is now Ankara. The region took their name: Galatia. They remained a distinct people for centuries, fighting, allying, being defeated by Rome in 189 BCE, becoming a client kingdom and then a Roman province in 25 BCE.

And here is the detail that makes it undeniable. Writing in the late fourth century CE — six hundred years after the migration — Saint Jerome observed that the Galatians spoke their own language, and that it was substantially the same as the speech of the Treveri of the Rhineland. A Celtic language, still spoken in central Anatolia, deep into the Christian era. The letter Paul addressed to the Galatians was addressed to them.

1071: the crossroads at Manzikert

The eleventh century is where the album sets itself, and the pivot is a single day in August 1071.

At Manzikert, near Lake Van in eastern Anatolia, the Byzantine emperor Romanos IV Diogenes met a Seljuk army under Alp Arslan. The Byzantine force was large, expensive and full of mercenaries — including Norman and Frankish contingents, some of whom left the field. Romanos was captured. The empire did not collapse that day, but the Anatolian plateau — the recruiting ground and the breadbasket — began slipping away, and within a generation the balance of the eastern Mediterranean had changed permanently.

Five years earlier, at the other end of the same road, the Normans had landed at Hastings.

That coincidence is the whole conceit of The Ancestry. Two ends of a continent, one generation, the same restless movement of armed men reordering who lives where and who remembers what. The instrumental track “Manzikert” sits at the centre of the album for that reason: it is the hinge the record turns on, and it is wordless because the thing it is about is not something anyone sang about at the time.

As I put it when the record came out: the memories that had passed into the resting places of history, myth, legend and forgotten happenings were the events that drew me to the stepping stone of remembrance and lineage — and that became the common thread holding the Celts together through the opening of this great cultural Pandora’s box.

What Byzantine and Celtic music actually share

Now the practical part, because a historical conceit does not make an arrangement work. Here is what the two musical systems genuinely have in common, and it is more than the romance suggests.

• Both are modal, not functional. Byzantine chant is organised by the oktoechos, a system of eight modes; Arabic and Ottoman music by maqam; Celtic traditional music by the church modes with pentatonic melodic habits. None of the three is built on functional harmony — the dominant-to-tonic engine that drives European art music from Bach onwards. In all three, the tonic is a place you depart from and return to, not a chord you resolve to.
• Both are drone-founded. Byzantine chant has the ison, a held vocal drone beneath the melody. Celtic music has pipes, hurdy-gurdy, open fiddle strings and the wire harp’s sustain. Melody-over-drone is the shared architecture, and it is the single most important reason the two combine without violence.
• Both treat ornament as structure. The Arabic and Turkish concept of ornament around a central pitch and the Irish cut, roll and cran are not the same vocabulary, but they occupy the same grammatical slot: they are how a line is articulated and how a performer signs it.
• Both are essentially monophonic in conception. One line, decorated, supported. Harmony, where it appears, is a modern addition in both worlds.

And the genuine incompatibility, which no amount of enthusiasm removes: intonation. Many maqamat use intervals that fall between the notes of equal temperament — the “neutral” second and third, sometimes loosely called quarter-tones. A lever harp cannot play them. A fretless oud or a ney can play nothing else comfortably. Ignore this and your arrangement will sound out of tune, and the listener will not know why.

Rhythm: where the fusion actually lives

The meeting point that works most reliably is not melodic at all. It is rhythmic.

Irish traditional dance music runs on a small number of metres: the reel in 4/4, the jig in 6/8, the slip jig in 9/8, the hornpipe in dotted 4/4. Turkish and Arabic music uses usul and iqa cycles, many of them additive: 9/8 divided 2+2+2+3 in the Turkish aksak family, 10/8 as samai thaqil, 7/8 in various groupings.

The slip jig and the aksak 9/8 contain the same nine pulses and divide them differently — 3+3+3 against 2+2+2+3. You can move between them without changing tempo, and the moment of transfer is one of the most genuinely thrilling things available in this idiom. It is also entirely defensible: nobody is being imitated, two rhythmic traditions are being placed in the same bar and allowed to share a pulse.

A century of people making this argument in sound

I did not invent this line of thought, and it would be dishonest to present it as though I had. It is worth setting out the lineage, partly out of fairness and partly because the failures in it are as instructive as the successes.

• The orientalist century (roughly 1850–1950). European composers wrote a great deal of music about the East using none of its actual materials — augmented seconds, a doumbek, and a set of assumptions. It is worth listening to precisely so you can recognise the reflex when it appears in your own writing. The tell is always the same: the imported element is scenery for a story told entirely in the importing tradition's grammar.
• The ethnomusicological turn (1950s onward). Field recording and serious study of maqam, raga and usul changed what was possible, because for the first time the systems were documented as systems rather than as colour. You cannot combine two grammars until you have learned that the second one is a grammar.
• The third-stream and world-fusion generation (1970s–80s). Musicians began working in mixed ensembles as peers rather than as employer and exotic hire. The recordings from this period that still hold up are, without exception, the ones where the players from each tradition were given authority over their own idiom.
• The Celtic eastward turn (1990s). Loreena McKennitt followed the historical migration route of the Celts across Europe and into Asia Minor and made records out of what she found — explicitly, and with the research in the liner notes. That combination of a documented premise and an honest statement of what was invented is the model I have tried to work from.
• The present. Cross-tradition collaboration is now ordinary, which raises rather than lowers the standard: when a tabla player is a phone call away, programming an approximation is a choice, and it is a worse one.

Rules I try to keep when writing across the road

Fusion is easy to do badly and the failure mode is always the same: a tradition used as decoration. These are the constraints I hold myself to.

1. One imported colour per texture. A duduk or an oud or a qanun. Three at once is not a fusion, it is a buffet, and the ear reads it as indecision.
2. Decide the tuning system before you write. If the piece is going to use neutral intervals, no lever harp and no piano can carry the melody — they can drone, and that is all. If the piece is equal-tempered, brief the oud player explicitly and do not ask for maqam inflections that will fight the harp.
3. Hire the tradition, do not impersonate it. If a piece needs a tabla part, hire a tabla player and ask for their idiom rather than programming an approximation. Credit them by name. This is both an ethical position and, plainly, better music.
4. Keep the drone honest. One drone instrument owns the sustained pitch. A Byzantine ison and a set of pipes in the same passage will drift against one another and the result is not ancient, it is out of tune.
5. Let the rhythm do the crossing. Metric modulation between a Celtic 9/8 and an aksak 9/8 is structural fusion. Adding a doumbek to a reel is costume.
6. Say what is invented. “Manzikert” is a new composition named after a real battle. It is not a reconstruction, there is no Byzantine source for it, and I would rather a listener know that than believe something untrue about the eleventh century.

Why the road is the right image

The most common objection to work like this is that it is anachronistic — that a bodhrán and a riq never met, that Galatian Celts and Irish harpers are separated by a thousand years and three thousand kilometres, and that putting them in one arrangement is a modern fantasy.

That objection is right about the facts and wrong about the conclusion. Of course this is a modern composition; every composition is. But the assumption underneath the objection — that musical traditions are sealed regional containers that only modernity has violated — is the one the history does not support. Tunes have always travelled with armies, pilgrims, sailors, refugees and traders. O’Carolan absorbed the Italian baroque within a generation of hearing it. The Breton revival absorbed the American folk boom in a decade. The traffic is the normal condition.

What the Galatian episode offers is not a licence to mix anything with anything. It is a specific, documented reminder that the Celtic-speaking world reached further east than the record-shop category suggests, and that the boundary between “Celtic” and “Middle Eastern” is a convenience of shelving rather than a fact of geography.

When I write from that position, the question stops being “may I combine these?” and becomes the far more useful “what did the road actually carry, and what am I adding?” The second question is answerable, and answering it honestly is most of the work.

Galatia Byzantium Manzikert maqam modal music fusion history The Ancestry