Mountains and the Sea: Landscape and History in the Ancient Greek World

Greece is a small peninsula at the southeastern tip of a large continent, almost but not entirely encircled by sea. That almost is everything. The Aegean is not a barrier separating land from land; it is a medium — a connected inland body of water whose islands are not obstacles but stepping-stones, rarely more than a day’s sail apart. Before any human society left a trace, the physical conditions for maritime connectivity were already in place.

The interior is a different story. The Pindus range runs like a spine from Epirus to the Peloponnese, rising to well over two thousand metres in places, splitting the peninsula into a mosaic of small plains, river valleys, and enclosed basins. No Greek plain is large. Boeotia and Thessaly are the nearest equivalents to breadbasket territory; Attica is modest; the plains of the Peloponnese are distributed across Argos, Laconia, and Messenia, separated by mountain passes. The terrain creates natural political units — small, defensible, prone to self-sufficiency — long before there are Greeks to inhabit them.

Climate reinforces this. The Mediterranean seasonal pattern — wet mild winters, hot dry summers — suits a particular agricultural system: olives, vines, grain in a rotation that can support modest population densities but demands labour-intensive management of terraced hillsides. This is an agriculture that binds people to specific microenvironments. The soil is thin on the slopes; the upland limestone karst drains fast; reliable water is scarce in summer. The result is a settled farming culture inherently fragmented into local communities, each tied to its own spring, its own valley, its own horizon of mountains.

Yet the sea pulls against all of this. The same climate that dries the interior in summer fills the Aegean with reliable northerly winds — the etesian winds — that make southward and eastward sailing predictable and northward sailing a calculated effort. Coastal and island communities are therefore not isolated but linked, if asymmetrically: sailing from Attica to Ionia is easier than the return passage. This asymmetry shapes migration and trade directions for millennia.

The terrain that fragments politics also structures diet. The Mediterranean triad — grain, vine, and olive — is not merely a cultural preference but a response to what the climate and soil actually permit. Olives require mild winters, heat in summer, and well-drained calcareous soils; they thrive on the terraced hillsides that are too steep and rocky for cereals, which is precisely why terrace-cultivation of olives makes ecological sense. Vines similarly tolerate drought and thin soils but are killed by hard frost. Both crops are therefore confined below roughly 500–700 metres altitude across most of mainland Greece, which means the highland interior of the Pindus and its subsidiary ranges is effectively olive-and-vine country only in the valleys. Cereal cultivation is constrained by the small size of the plains; even in Boeotia and Thessaly, the most grain-productive regions, the area under cultivation was modest by Near Eastern standards.

The highlands do something else entirely. Above the winter snowline, the terrain of Epirus, Thessaly, and the Peloponnesian uplands supports pastoral economies: sheep and goat herding on summer mountain pastures with seasonal movement to lowland winter grazing — the practice known as transhumance. This is not a marginal activity: it connects lowland grain-farming communities with highland pastoralists through networks of exchange in cheese, wool, hides, and animals for sacrifice. The mountain ranges that appear to fragment the Greek world also sustain the animal economies that supplement its grain-based agriculture. The polis relies on its herdsmen no less than its ploughmen, and the herdsmen live in the landscape that the mountains make.

The sea adds the third component. The Aegean is not only a trade corridor; it is a food system. Tuna pass through the Aegean and Black Sea straits on their annual migration, and their predictability made them a dietary staple at coastal sites from the Mesolithic onward — the fishing assemblage at Franchthi Cave in the Argolid, documented as early as 11,000 BCE, includes open-sea tuna that required boats to catch. Shellfish, sea urchin, and smaller fish supplemented the diet at coastal and island sites throughout the prehistoric period. The implication is significant: the peoples of the Aegean coast and islands ate a different diet from those of the interior, one with substantially more marine protein, and this dietary difference is now, as discussed in the palaeodiet section below, directly recoverable from skeletal isotope evidence.

Transhumance — the seasonal vertical movement of herders and their flocks between lowland winter pastures and high summer pastures — is one of the most geographically determined practices in the Greek world, and one of the most temporally persistent. The same mountain corridors used by the Vlachs and Sarakatsani herders of the nineteenth and twentieth centuries, documented in detail by Wace and Thompson in 1913, follow routes whose archaeological signatures reach back to the Neolithic and possibly earlier. This is not a coincidence: the routes are determined by topography. The Pindus chain and its subsidiary ranges offer summer grazing above the snowline from roughly May to October; the plains of Thessaly, Boeotia, and the Epirote coastal lowlands offer winter pasture. The gradient between them is not optional — it is what the terrain imposes on anyone who herds sheep and goats in this landscape.

The question of when organised transhumance began is genuinely debated. Archaeological surveys of the Grevena highlands in the north-eastern Pindus — the watershed between western Macedonia and Epirus, centred on the modern village of Samarina — have revealed seasonal occupation sites from the Middle Palaeolithic through to the Bronze Age, suggesting that the high-altitude routes were in use by hunter-gatherers and later by pastoral communities long before the classical period. Whether this constitutes true transhumance in the agricultural sense, or simply seasonal use of upland hunting territories, remains open. What is not in doubt is that by the historical period the practice was sufficiently entrenched that Greek sources take it for granted rather than describing it — a reliable sign that something is structural rather than exceptional.

Isotopic analysis is now providing direct evidence. Stable isotope studies of sheep and goat tooth enamel from Classical and Hellenistic sites in Thessaly — the region where the transhumance debate has been most active — show both sedentary and seasonally mobile patterns of animal management at the same sites, sometimes within the same assemblage. This does not mean that all herders were transhumant; it means that different management strategies coexisted in the same landscape, probably reflecting different household strategies, different herd sizes, and different relationships to the urban markets that were beginning to demand reliable meat and wool supplies at scale.

Transhumance is structurally significant for this document because it is a form of landscape use that operates outside — and in some respects against — the logic of the palace economy and the polis. Palace redistribution systems depend on surplus extraction from sedentary agricultural producers; the polis depends on a cultivated chora with a defined citizen body. Transhumant communities do neither: they move seasonally across territorial boundaries, they do not till fixed fields, and they were rarely fully integrated into the civic structures of the poleis whose uplands they used. In the Greek sources they appear as marginal, sometimes threatening — the Arcadians with their acorn-eating reputation, the Epirote pastoralists who in the historical period remain organised as tribal confederacies rather than city-states long after the rest of Greece has urbanised. The mountains explain this directly: where the terrain cannot sustain an agricultural chora dense enough to support a polis, the polis does not form. What forms instead is a pastoral society, seasonally mobile, genealogically organised, and in a complex, often tense relationship with the grain-and-olive world on its borders.

The persistence of transhumance into modernity — it was still an economically significant practice in northern and central Greece into the mid-twentieth century — is itself an argument about the power of geography. The Ottoman state, the Greek national state, and the European economic order all tried with limited success to sedentarise transhumant communities. The mountain topography that drove the practice in antiquity continued to make seasonal movement the most rational response to the terrain. This is one of the clearest examples in the Greek world of a practice genuinely determined by geography — not in the sense that individual herders had no choices, but in the sense that the structure of the practice recurs whenever the combination of high mountain summer pasture and accessible lowland winter pasture is present, regardless of the political system governing the territory between them.

The shepherd’s dog is one of the most geographically persistent elements of the Greek pastoral world — and one of the clearest links between the transhumance evidence discussed above and the deeper prehistory of the human relationship with animals. Dogs were the first domesticated species and the only large carnivore to enter a domestic relationship with humans during the Pleistocene. Their domestication from wolves predates agriculture by millennia and is therefore inseparable from the hunter-gatherer context: the earliest domestic dogs were hunting companions and camp guards, not herding or guarding tools. That functional transition — from hunting partner to livestock guardian — mirrors the agricultural revolution itself.

The earliest confirmed domestic dog is the Bonn-Oberkassel specimen from Germany, dated to approximately 14,000 BCE. It was buried alongside two humans with grave goods, and analysis of its teeth shows it had been kept alive for weeks while gravely ill with canine distemper — evidence of emotional attachment, not merely utilitarian use. The dog was already a companion before it was a worker. This co-burial is itself geographically structured: the dog travelled with a mobile hunter-gatherer community whose seasonal movements through the landscape are the deep precedent for the transhumance routes that later Greek herders would use.

When the Neolithic transition occurred, the dog’s functional repertoire split. Some dogs retained their original hunting role — the Greek Laconian hound, described in detail by Xenophon in his Cynegeticus (c. 360 BCE), is a swift scent-tracking hunting dog descended from this original function. Others were repurposed as flock guardians: where humans had once kept dogs to help catch animals, they now kept dogs to prevent other predators from taking the animals they herded. The Molossian of Epirus represents this second lineage. The two functional types that Aristotle distinguishes in his Historia Animalium (c. 350 BCE) — hunting dog and sheep-dog — are therefore not simply different breeds but the archaeological record of two different phases of the human–dog relationship, both present simultaneously in the historical Greek world.

Numbers and management practice are not well documented for the Greek period specifically, but Roman agronomists provide the nearest comparative evidence. Varro (Rerum Rusticarum II.9, c. 37 BCE) discusses flock-guardian dogs at length, recommending white-coated animals so shepherds can distinguish them from wolves at night, advising spiked collars as throat-protection, and noting that the number of dogs should be proportional to flock size and wolf pressure. Columella (De Re Rustica VII.12, 1st c. CE) recommends an odd number of dogs per flock to avoid deadlock when the pack divides to confront a threat from multiple directions. Both authors are drawing on a continuous tradition of pastoral dog management that extends without structural break from the Greek period. The Linear B tablets name shepherds and count their flocks precisely but do not name individual dogs — the dogs are infrastructure, assumed rather than enumerated.

The deeper point, for this document’s argument, is that the shepherd’s dog participates in transhumance directly. Livestock guardian dogs are not kennel animals; they live permanently with the flock, sleeping among the sheep, eating what the shepherd eats, and travelling every transhumant route. The Greek Shepherd breed is today specifically documented as crossing national borders with its flocks on annual migrations. This is a form of geographical determination operating through biology: the mountain topography that demands transhumance also demands a specific category of dog, and that demand has maintained functionally similar animals in these landscapes for millennia. The dog, like the route it walks, is part of what the Pindus imposes.

The Early Bronze Age Aegean is not yet a world of palace economies or literate bureaucracies. What it is, unmistakably, is a maritime network — and that network is substantially driven by geography. The Cycladic islands, scattered across the central Aegean like the stepping-stones they literally are, emerge in this period as production and distribution nodes rather than mere waypoints. Their specific resources — the obsidian of Melos, the marble of Paros and Naxos, the emery of Naxos — are not accidents of culture. They are gifts of geology that become the basis of exchange.

Melos is the clearest case. Its volcanic obsidian is the finest knapping stone in the Aegean basin, and its distribution in the archaeological record reaches back to the Mesolithic. By the Early Bronze Age, obsidian from Melos appears at sites across the Aegean and into Anatolia. This is not evidence of a Melian state conducting trade missions; it is evidence that the physical properties of one island’s geology made it a node in networks that would have formed regardless of who lived there. The island’s position — roughly central in the Cycladic group, with reliable anchorages — reinforced the effect.

Cyprus enters the eastern horizon of these networks by the mid-third millennium, initially as a supplier of copper from the Troodos massif. The island’s location between the Aegean world and the Levantine coast makes it simultaneously peripheral to each and central to both as a bridge. This dual peripherality — central only by virtue of being between others — is a recurring geographical role in Mediterranean history, one that Cyprus will play continuously for two millennia.

The mainland at this stage is less conspicuous in the archaeological record than the islands and the Anatolian coast. The terrain that fragments the mainland into small polities in the historical period also fragments the Early Bronze Age archaeological record into regional cultures — Argolid, Corinthia, Boeotia — that share broad similarities but lack a unifying material signature as distinctive as the Cycladic.

The Late Bronze Age is the moment when the Aegean world becomes legible in the wider eastern Mediterranean record. The Amarna Letters — the royal diplomatic archive of fourteenth-century Egypt — mention Alashiya (Cyprus), the Ahhiyawa (almost certainly a Mycenaean Greek kingdom), and the great powers of Hatti, Mitanni, Assyria, and Babylon. The Aegean has been connected to this diplomatic world for some time before the archive is written, but the archive makes the connection visible.

What the geography explains is not membership in this system but position within it. Cyprus — Alashiya — holds a position no Aegean island can replicate: large enough to support a complex palace economy, rich enough in copper to be indispensable to the bronze-dependent militaries of Egypt and Hatti, and positioned exactly at the junction of eastern Mediterranean sea routes. Ugarit, on the Syrian coast, plays the complementary role of transit hub: a port city whose prosperity rests not on local resources but on geographical position as the place where Aegean, Cypriot, Egyptian, and Levantine goods are exchanged.

The Mycenaean palaces of the Greek mainland — Mycenae, Tiryns, Pylos, Thebes — are embedded in this system in ways their Linear B tablets confirm. The tablets record enormous quantities of bronze weapons and tools, suggesting that copper supply was a strategic concern. The geography of the Argolid is not large enough to feed a great empire. Mycenaean power is maritime and extractive: it reaches into the Aegean island world and across to Anatolia and Cyprus for the materials its palace economy requires.

Thera (modern Santorini) interrupts this story in a way that has never been fully resolved. The Minoan-period eruption — one of the largest volcanic events of the Holocene — destroyed the island’s northern inhabited area and deposited ash across the eastern Mediterranean. The dating debate (late seventeenth or mid-fifteenth century BCE) determines whether the eruption caused or coincided with the decline of Minoan palatial culture on Crete. The geography is clear; the historical interpretation remains contested.

The Late Bronze Age collapse around 1200 BCE stripped away the institutional superstructure — palace bureaucracies, centralised redistribution networks, scribal culture — that had channelled and recorded geographical connections. The connections themselves, at a lower intensity and smaller scale, persisted. The mountains remained between the plains. The Aegean remained navigable in summer. The islands remained where they had always been.

Lefkandi on Euboea is the key site for understanding this persistence. In the tenth century BCE, at the apparent nadir of the Greek Dark Ages, Lefkandi was importing Cypriot bronze, Levantine faience, and Egyptian gold. Lefkandi’s geographical advantage is clear: Euboea flanks the main channel between central Greece and the northern Aegean; its position on the Euripus strait places it astride the coastal route between Thessaly and Attica. This is a location that traffic must pass, regardless of which political superstructure is or is not operating.

The broader pattern of the Dark Ages is one of differentiated contraction rather than uniform collapse. Communities on or near the main maritime routes — Euboea, the eastern Aegean coast, Cyprus — maintained levels of exchange connection that communities in the interior or on geographically marginal routes could not. This is strong evidence for the persistence of geographical constraints even through institutional discontinuity.

With the collapse of palace redistribution networks, diet would have contracted sharply for most of the Greek population. The evidence from skeletal assemblages of this period, where it exists, reflects this: a narrowing of the dietary range, greater reliance on locally available staples, and a reduction in the markers of dietary diversity that better-connected Bronze Age populations show. The palaeodiet section below discusses how isotopic science recovers this kind of information from the bones themselves.

Between roughly 800 and 500 BCE, Greek communities founded colonies — apoikiai — around the Mediterranean and Black Sea coastlines with a geographic logic that is unmistakable in retrospect. The destinations cluster at river mouths giving access to agricultural hinterlands, natural harbours on routes between the Aegean and the western Mediterranean, and positions controlling approaches to metal-rich regions. Colonisation was geographically structured by the same logic that had structured Bronze Age exchange: uneven distribution of resources, manageable crossing distances, the advantage of position.

The Euboeans were first, and the Euboean cities were the heirs of Lefkandi’s Dark Age connectivity. Pithekoussai on the island of Ischia — the earliest Greek colonial foundation in the west, c. 770 BCE — is positioned to access Etruscan metals without confronting the Etruscan coast directly. The Black Sea colonisation, dominated by Miletus, follows a different logic: the Bosphorus and Hellespont are navigational chokepoints beyond which lies the Pontic steppe, a source of grain, fish, timber, and slaves on a scale the Aegean could not match.

Within Greece itself, the polis world that emerges in the eighth and seventh centuries is shaped by the same terrain that had created the Bronze Age mosaic. Each polis characteristically controls a defined agricultural territory — a chora — bounded by mountains or sea. The chora is not just a political concept; it is a nutritional one. The polis needs its hinterland to feed its citizens, and the size of the hinterland that the terrain makes available determines, in the first instance, how large the polis can grow. Athens, with thin soils in Attica, must import grain from the Black Sea — which is why the silver of Laurion and the fleet it funds are existential assets, not optional luxuries. Corinth, astride the Isthmus, controls the flow of agricultural produce between east and west Greece as much as the flow of manufactured goods.

By 479 BCE the Greek world had spread from Massalia to Phasis on the eastern Black Sea. The geography of the expanded Greek world is not a random scatter but a map of navigable routes and resource concentrations — one that Cunliffe’s framework of maritime corridors as the primary connectors of early European history helps to read.

Fish is the most geographically specific food in the Greek world. Every other staple — grain, oil, wine, meat — could be produced inland, stored, and taxed at the city gate. Fish required the sea, or at minimum a lake or major river, and the sea was, as the scholarship discussed below makes plain, structurally resistant to the kind of territorial control that the polis exercised over its agricultural hinterland. The history of fish in the Greek world is therefore simultaneously a history of diet, social status, economic regulation, and the limits of civic authority. It is also, at a deeper level, a history of the sea’s meaning: a medium that was food source and disposal ground, highway and lair of monsters, simultaneously.

The dietary trajectory runs from Mesolithic subsistence to classical luxury with a complicated middle. At Franchthi Cave, open-sea tuna fishing was practiced from at least 11,000 BCE; marine protein was a dietary staple for coastal Aegean communities throughout the prehistoric period. The Bronze Age seascape evidence, discussed below, shows fishing and shellfish-gathering as embedded in the daily and ritual life of coastal communities. But by the classical period something had changed: fish had acquired a second identity as a luxury item, morally loaded and politically charged, quite independent of its dietary function. Fresh fish — above all eel from Lake Copais, tuna from the Black Sea, and various prestige species marketed through the Athenian agora — became the marker of a particular kind of appetite, associated with lack of self-control, democratic excess, and the dangerous pleasures of the marketplace.

The ambiguity that Berg identifies in the Bronze Age material reaches its fullest articulation in the classical period through a different route: the philosopher, the orator, and the comic playwright. James Davidson’s 1993 article in the Classical Quarterly demonstrates that references to fish consumption in classical Athens are remarkably frequent and remarkably loaded. The word opsophagos — “relish-glutton,” specifically a compulsive fish-eater — is a term of social and moral condemnation: Demosthenes deploys it against Philocrates, Aeschines against Timarchus, and comic writers use it repeatedly. The connection between fish-eating, sexual incontinence, and political untrustworthiness is not coincidental: it reflects a coherent ideology of appetite and self-control at the heart of democratic Athenian citizen identity.

The key structural reason for this, developed further in Davidson’s 1997 book, is that meat was sacrificed and its portions distributed equally among citizens at civic festivals — eating sacrificial meat was an act of communal participation that could not be perverted by individual appetite or purchasing power. Fish, by contrast, was not used in sacrifice. It was bought and sold in the fish market, and you could buy as much as you could afford, or more than you could afford. Fish-buying thus indexed individual appetite in a way that meat-eating could not, and excessive fish-buying was a legible sign of the kind of ungoverned desire that democratic ideology regarded as a threat to civic order.

The Hinterland dimension connects directly to this document’s core argument. Jennifer Moody’s 2012 study in British School at Athens Studies examines Bronze Age and Iron Age Crete through the framework of “hinterlands and hinterseas” — the productive zones, both terrestrial and maritime, that surrounded and supported the major centres. The hintersea concept is important: the sea is not simply a route between settlements but a productive zone in its own right, yielding protein, dye-stuffs, and trade goods, and sustaining communities whose economic logic was oriented outward toward the water rather than inward toward the agricultural plain. Chryssi Island, documented in Mylona’s study, is precisely such a hintersea resource zone: a small island with no permanent agricultural community, used seasonally for specialised marine production.