Derinkuyu and the Underground Cities of Cappadocia Astronomical Alignments

About Derinkuyu and the Underground Cities of Cappadocia Astronomical Alignments

In 1963 a homeowner in the Anatolian town of Derinkuyu knocked out the back wall of a cellar and stepped onto the rim of a shaft that descended eighty-five meters into the bedrock. The rooms opening off that shaft turned out to run eighteen levels deep, large enough for perhaps twenty thousand people, and pierced top to bottom by narrow vertical flues that open directly onto the sky. Yet no published archaeoastronomical survey has demonstrated a deliberate astronomical alignment at Derinkuyu to the precision standard applied at Stonehenge, Newgrange, or Chichen Itza. The underground cities of Cappadocia were dug into tuff, not raised toward the sky, and the absence of a coherent alignment literature for them is itself the honest starting point. What exists is weaker, more ambiguous, and more interesting than a confirmed solstice axis: a set of vertical shafts that could serve as zenith windows, a regional context saturated with Hittite solar theology, and a surface landscape whose volcanic pillars mark horizon positions that any attentive observer would read as a calendar. The page that follows walks each of those threads carefully, names the researchers and publications that back every load-bearing claim, and marks clearly where evidence is thin or unpublished.

A search of the Journal for the History of Astronomy, Archaeoastronomy: The Journal of Astronomy in Culture, and the proceedings of the SEAC (Société Européenne pour l'Astronomie dans la Culture) conferences through 2025 turns up no peer-reviewed archaeoastronomical survey of Derinkuyu or the other major Cappadocian underground cities — Kaymaklı, Özkonak, Mazı, Tatlarin. This is in sharp contrast to the dense alignment literature that exists for Anatolian surface monuments of the same broad periods. A. César González-García and Juan Antonio Belmonte, writing in Clive Ruggles's Handbook of Archaeoastronomy and Ethnoastronomy (Springer, 2015), surveyed a statistically significant sample of Hittite temples and monumental gates and demonstrated that their orientations are not randomly distributed but cluster in patterns interpretable within Hittite religious culture. Cappadocian underground sites, however, have not been surveyed by that team or by any other research group that publishes in the field. The Handbook itself indexes no entry for Derinkuyu. The English-language popular claim that the shafts "align with Sirius" or "frame Polaris" cannot be traced to any primary archaeoastronomical measurement; it circulates through travel writing and speculative literature without a citable source.

The responsible reading of this silence: it is premature to claim a confirmed alignment, and equally premature to rule one out. The sites are difficult to instrument (access to the deepest levels at Derinkuyu is restricted, and the shafts at depth are too narrow to carry standard survey equipment), and the political and funding environment for open-access archaeoastronomy in central Turkey has not prioritized subterranean sites. A rigorous survey would require laser rangefinding up each ventilation shaft with corrections for atmospheric refraction, dated coring of the shaft walls to establish construction-phase chronology, and comparative measurement across the whole network of two hundred-plus underground complexes. That work has not been done.

What can be said with confidence is that the geometry of a narrow vertical shaft viewed from a deep chamber functions as a collimating tube. This is not a speculative claim — it is an optical fact. A shaft of diameter d and depth h restricts the visible sky to a cone of half-angle arctan(d/2h). Derinkuyu is pierced by more than fifteen thousand narrow auxiliary flues of roughly ten to twenty-five centimeters diameter and by a smaller number of wider central shafts that double as wells. Taking one of the main central shafts — roughly one meter in diameter and sixty meters deep — the visible-sky cone from the bottom is about one degree of arc, roughly twice the angular diameter of the full moon. For a narrow auxiliary flue of fifteen centimeters diameter and the same depth, the cone collapses to about a seventh of a degree. Any chamber positioned below such a shaft is, in effect, a zenith tube; which shaft type you measure changes the angular tolerance but not the geometric principle.

The archaeoastronomical literature on deliberately built zenith tubes is substantial. Anthony Aveni's Skywatchers (revised University of Texas Press edition, 2001) describes the zenith tubes at Monte Albán and Xochicalco, where narrow vertical apertures mark the exact dates the sun stands directly overhead at the tropical latitude of the site. Xochicalco's tube sits at 18.80°N, where the noon sun shines straight down the shaft on May 14–15 and July 28–29 with no shadow on the walls. Derinkuyu sits at latitude 38.37°N, well north of the Tropic of Cancer, so the sun never reaches zenith there — a solar zenith tube is geometrically impossible at that latitude. But stellar zenith tubes remain functional: any star with a declination close to +38.37° passes directly overhead once per sidereal day, and the question of which stars occupied that declination band during Derinkuyu's probable expansion (roughly 800 BCE to 700 CE, depending on whose dating one accepts) is answerable from precession tables. Rather than stake the argument on any single named star — precession moves the whole grid, and the bright stars that sit near declination +38° today (Vega, at +38.78°) were considerably higher in the sky two and three thousand years ago — the geometric point is the one that holds: any star whose declination matched the site's latitude during the period of use would transit at zenith for an observer looking up the shaft.

The Egyptian comparison — made often in popular literature, including Robert Bauval and Adrian Gilbert's The Orion Mystery (Crown, 1994) — proposes that the so-called star shafts of the Great Pyramid were designed to frame specific stars at meridian transit during Khufu's reign. The Bauval/Gilbert hypothesis has been contested on archaeological grounds by Kate Spence, Mark Lehner, and Zahi Hawass, and on astronomical grounds by Ed Krupp (Griffith Observatory) and Giulio Magli. The mainstream Egyptological view treats the shafts as religious features, possibly also serving ventilation, and holds that their specific alignments, while real to within a degree or so, were chosen from a menu of near-meridian stars rather than designed around any single stellar target. The same cautionary frame applies, and with greater force, to Derinkuyu: absent a systematic survey of what the shafts in fact frame, any specific stellar claim is a guess.

The cultural environment in which Derinkuyu was used includes one of the best-documented systematic astronomies of the ancient Near East. Hittite ritual texts, preserved on cuneiform tablets from the archives of Hattuša (near modern Boğazkale, about 180 kilometers north of Derinkuyu), reference the Sun Goddess of Arinna as a primary deity and record ceremonial observations timed to solar and lunar cycles. González-García and Belmonte's survey of Hittite monumental orientations, summarized in the Handbook of Archaeoastronomy and Ethnoastronomy (Springer, 2015), examined a statistically significant sample of Hittite temples and monumental gates and found that their orientations cluster non-randomly at declinations corresponding to the solstices and equinoxes. Temple 1 at Hattuša itself shows an axial orientation toward the winter-solstice dawn, with the sun rising above the northern rim of the acropolis where a subsidiary shrine of the goddess was located.

The rock-cut sanctuary of Yazılıkaya, a short walk from Hattuša, extends this pattern. Studies by Eberhard Zangger and Rita Gautschy (published in Journal of Skyscape Archaeology, 2019) identify the complex as a calendrical device, with the two main chambers reading as structured representations of the solar year and the lunar month; the gatehouse of Chamber A admits the sun on specific days of the year so that it falls on a marked figure. Whatever the exact chronology of Derinkuyu's deepest levels (an open question, with conservative dates placing first major excavation in the Hittite period of the second millennium BCE and other dates as late as Byzantine), the successive cultures that dug and used these underground cities inherited a documented Anatolian solar and stellar tradition. It would be anomalous if that tradition left no trace underground.

The Phrygian and Byzantine-era inhabitants of the Cappadocian sites belonged to different cultural and religious frames, each with its own astronomical and calendrical machinery. The Phrygians (roughly the twelfth through seventh centuries BCE as a kingdom, with cultural continuity for centuries beyond the Cimmerian destruction of Gordion around 696 BCE) worshipped Matar (Cybele) and Sabazios, and their rock-cut monuments — Midas City, the tomb of Midas, the cult facades at Aslankaya — are oriented with a clear preference for cardinal directions, especially east. The Byzantine Christian community that used the upper levels of Derinkuyu for refuge between the 7th and 13th centuries CE brought the Julian liturgical calendar and the east-facing orientation of Christian church architecture, visible in the rock-cut churches of the Göreme valley a short distance north.

The Cappadocian volcanic landscape provides a built-in horizon calendar for any community using it attentively. The fairy chimneys — tall narrow pillars of tuff capped by harder basalt, some reaching 40 meters in height — create a pattern of sharp vertical silhouettes against the eastern and western skylines. Erciyes Dağı, the dormant stratovolcano 3,917 meters high and visible from virtually every underground-city entrance in the region, dominates the eastern horizon. A community watching the sun or a bright star rise against that horizon would track its annual drift between specific named chimneys or against the shoulder of the mountain, generating a practical calendar without any constructed instrument.

Horizon calendrics of this kind — using natural or minimally modified landmarks rather than purpose-built sight-lines — is the subject of extensive ethnographic literature. Stephen McCluskey's Astronomies and Cultures in Early Medieval Europe (Cambridge, 1998) documents the practice in European monastic communities. Ray Williamson's Living the Sky (University of Oklahoma Press, 1984) documents it across the American Southwest. The pattern is close to universal wherever agricultural communities live with a strongly marked natural horizon. For the Cappadocian underground city builders, the external landscape did not need to be augmented with stone circles or standing stones to serve as a calendrical instrument. The tuff pillars were already in place.

The strongest skeptical position, held by most mainstream archaeologists of the region, is that the ventilation shafts at Derinkuyu are ventilation shafts. The need for airflow in a multi-level underground settlement is absolute — without forced or passive convection, the lowest chambers rapidly become uninhabitable — and the engineering solution arrived at (a central deep shaft paired with angled auxiliaries, using differential temperature and pressure to drive convection) is elegant and sufficient. No astronomical secondary function is required to explain the shafts' existence, their depth, or their number. Occam's razor cuts against elaborate alignment hypotheses in exactly this situation.

A weaker but still serious counter-argument addresses the zenith-tube claim specifically: even if an individual shaft happens to frame a star at meridian transit, this is not by itself evidence of intention. In any large random set of vertical apertures, some will, by chance, frame bright stars. The null hypothesis — chance framing — has not been tested against the Derinkuyu shaft data because the data have not been systematically gathered. Until someone measures every shaft's azimuth, declination, and framed arc, and runs a statistical test against the expected distribution for randomly oriented apertures, no alignment claim can be distinguished from noise.

The alternative explanation that deserves the most attention is that the underground cities were not primarily astronomical environments at all — they were refugia, storage systems, and in later periods Christian monastic habitations. The question of alignment, asked of a site whose architectural logic is dominated by defense and sustained habitation under siege, may simply be the wrong question. The more productive question is what it meant to live without sky access for weeks or months at a time. That is a question about sensory deprivation, ritual isolation, and the maintenance of calendrical awareness by other means — and it opens onto a rich comparative literature that the alignment question does not.

A population sheltering in Derinkuyu during an extended siege or persecution — the probable primary use during the Byzantine period — lost the ordinary sensory apparatus by which human communities track time. No sunrise and sunset. No moon. No stars. Temperature stable year-round near 13°C. Acoustic environment muffled and resonant rather than open. Light source: olive-oil or tallow lamps. This environment has parallels in the kiva tradition of the Ancestral Puebloan Southwest (where sky access was controlled through a single roof hatch and a ladder), in the Tibetan Buddhist dark retreat (yangti) practice described in teachings by Lopön Tenzin Namdak and in Chögyal Namkhai Norbu's Bon and Dzogchen lineages, and in the subterranean incubation chambers beneath Greek oracle sites at Trophonius (described in Pausanias, Description of Greece 9.39) and Eleusis. Each of these traditions developed internal calendrical and contemplative technologies adapted to prolonged sky-deprivation.

Whether the Cappadocian community developed analogous practices is undocumented — the population that used the deepest levels left no written trace there — but the comparative record suggests it is plausible. The maintenance of a liturgical calendar during extended refuge would have required either a runner to the surface, a designated window-observer at a ventilation shaft, or an internal time-keeping technology such as a candle clock or graduated oil lamp. The small unventilated chambers at the deepest levels, too confined for ordinary habitation and lacking obvious storage or defensive function, are the likeliest candidates for ritual isolation or vision-seeking spaces in the kiva or yangti mold. This remains a hypothesis; no excavation has yielded the kind of residue (ceremonial vessels, carved wall markings, isolated human burials) that would confirm a ritual interpretation over a purely utilitarian one.

Malta's Hal Saflieni Hypogeum (roughly 4000 to 2500 BCE) has stronger astronomical alignment claims than Derinkuyu. Heritage Malta and UNESCO documentation of the so-called Holy of Holies chamber report a winter-solstice light-shaft alignment in which the rising solstice sun penetrates the chamber through a specific aperture. The Hal Saflieni case is instructive because it shows what a defensible underground alignment claim looks like: a specific measured azimuth, a dateable architectural feature, and a solar event whose geometry matches within the tolerance of the measurement. Derinkuyu does not currently meet this standard; Hal Saflieni does.

Newgrange (County Meath, Ireland, roughly 3200 BCE) is the cleanest case of a deliberate subterranean-to-sky alignment: a constructed roof-box admits the rising sun into the central chamber for roughly 17 minutes on and around the winter solstice, a feature demonstrated by Michael O'Kelly's 1967 observation of the winter-solstice event and subsequent excavation reports (1962–1975, published in his Newgrange: Archaeology, Art and Legend, Thames & Hudson, 1982). The functional logic at Newgrange is the inverse of a zenith tube — the light comes in horizontally at sunrise rather than vertically at meridian transit — and it shows how a monument can be both underground and astronomically aligned when the builders intend it. No equivalent feature has been documented at Derinkuyu.

A serious archaeoastronomical survey of the Cappadocian underground cities is overdue. The instruments needed (laser rangefinders, precision digital compasses, modern atmospheric-refraction correction tables) are standard and inexpensive. The intellectual question — whether any of the 200-plus underground complexes in the region show deliberate astronomical design, whether the shafts frame specific stars or the meridian, whether solar or lunar penetration of the entry corridors occurred on significant dates — is answerable with field work that has simply not been done. Until it is, honest writing on the subject names what is known, what is inferred, and what is guessed, and leaves the final verdict where it belongs: with the measurements, when someone finally takes them.

Several adjacent questions deserve attention from the same survey program. The entry-corridor azimuths at Derinkuyu and at neighboring Kaymaklı have never been measured against the positions of the solstitial sun or the full moon at the major lunar standstill; an orientation survey of the tunnel entrances would produce a small, cleanly testable dataset. The tuff bedrock of the Cappadocian plateau preserves chisel marks cleanly, and any astronomical inscription — sun-wheel glyphs, star-groupings, calendrical notation — that survives on the shaft walls could in principle be dated by context. The acoustic properties of the largest communal chambers, with their consistent volumes and vaulted tuff ceilings, lie close to the resonance range documented at Malta's Hal Saflieni Hypogeum by Debertolis, Coimbra, and Eneix in 2014, and a comparative acoustic survey would be inexpensive and revealing. None of this has been done. The gap between what could be known and what is currently known is wide, and the answer to that gap is the slow empirical work of measurement, not further speculation at a distance from the site.

Significance

The question of astronomical alignment at Derinkuyu matters beyond the narrow issue of where the shafts happen to point. It is a test case for how archaeoastronomical claims should be made and evaluated. The gap between confident popular claims and absent peer-reviewed studies is wider at Derinkuyu than at almost any other famous ancient site, and that gap is a useful teaching moment for any reader learning how to read archaeological and astronomical literature critically.

For the history of Anatolian astronomy, Derinkuyu's unsurveyed status is a genuine loss. The region produced, across several successive cultures, a documented observational tradition — Hittite solar theology, Phrygian mountain cults, Byzantine monastic astronomy — and the underground cities sit geographically at the center of that tradition. If the Cappadocian shafts do frame specific stars at meridian transit, or if the entry corridors admit the solstitial or equinoctial sun, this would provide material evidence for underground astronomical practice that currently exists nowhere else in the Anatolian record. If they do not, the absence is itself informative: it would suggest that the astronomical functions of the Hittite surface temples did not transfer to the subterranean refuge architecture, and that the underground cities were conceived by their builders as working environments disconnected from the celestial calendar of the surface world.

For Satyori's broader framework on ancient technology and inherited knowledge, the Cappadocian case illustrates a principle that runs through the whole field: the survival of a physical structure does not entail the survival of its original meaning. The modern residents of Derinkuyu used the upper chambers for centuries as cold storage and wine cellars without understanding that the tunnels extended eighty-five meters below their houses. The 1963 discovery of the lower levels surprised the community living directly above them. This pattern — active physical use combined with total loss of original context — is the rule rather than the exception for very old structures. What was clear to the original builders about the purpose of every shaft, chamber, and corridor is now a scholarly question to be answered by measurement and comparison, with no possibility of appeal to living tradition.

The honest answer to the alignment question — "we do not know, because no one has done the survey" — marks the edge of what archaeoastronomy has so far accomplished in this region and identifies a concrete research program that could be carried out inexpensively and would yield useful results whether the outcome is positive or negative. Sites that sit in this epistemic position deserve careful handling: neither dismissed because current evidence is thin, nor over-claimed because the speculative possibilities are dramatic. The discipline of holding that middle position is part of what it means to work seriously with ancient monuments.

The broader cultural significance of the Cappadocian underground network — the sheer scale of the engineering, the longevity of the occupation across cultures and religions, the evidence it provides for civilizational capacity during the long pre-modern periods — is largely independent of the alignment question. These are real sites with real technological achievements regardless of whether any shaft happens to point at Sirius. Allowing the alignment question to remain open, and locating it properly as a specific empirical problem awaiting fieldwork, protects the site's actual significance from inflation by speculative overlay and from dismissal by reactive skepticism.

Connections

Derinkuyu's alignment status is best understood in comparison to other underground sites with stronger or weaker astronomical claims. The cleanest comparative case is Newgrange, where a measured winter-solstice penetration of the central chamber was demonstrated by Michael O'Kelly in 1967 — a standard the Cappadocian sites cannot currently match. See the sister article on Newgrange's solstice alignment for the contrast.

The Maltese underground ceremonial architecture at Hal Saflieni Hypogeum sits at the intermediate position: Heritage Malta documentation reports a winter-solstice light-shaft alignment in the so-called Holy of Holies chamber, a claim Derinkuyu's shafts have not been rigorously tested for. If the Cappadocian sites were surveyed with the same method, they would either join Hal Saflieni as demonstrated subterranean calendrical sites or be established as astronomically null, and either finding would be informative.

The regional context places Derinkuyu within the broader Anatolian archaeoastronomical record. The Hittite capital at Hattuša (Boğazkale), with its documented winter-solstice dawn alignment of Temple 1 and the calendrical geometry of the nearby rock sanctuary at Yazılıkaya, provides the cultural baseline for what astronomical observation looked like in central Anatolia during the second millennium BCE. The earlier and more dramatic monumental construction at Göbekli Tepe (c. 9500 BCE) extends the question of ancient astronomy in the region back by six or seven thousand years, and recent work by Martin Sweatman and Dimitrios Tsikritsis on possible constellational references in the Göbekli Tepe reliefs has been met with both serious interest and serious critique.

The comparison with Egyptian subterranean architecture is unavoidable. The Great Pyramid's so-called star shafts, the subject of Robert Bauval's stellar-correlation hypothesis and Kate Spence's counter-argument, present the same structural question as Derinkuyu's ventilation shafts — does a narrow vertical or angled aperture framing a specific sky patch constitute evidence of astronomical intention? — and the same evidentiary difficulty. The Egyptian case has been surveyed in detail; the Cappadocian case has not.

The ritual-use comparison draws on a different family of sites. The kivas of the Ancestral Puebloan Southwest, the dark-retreat (yangti) caves of the Tibetan Buddhist tradition, the incubation chambers at Eleusis and at the Trophonius oracle in Lebadeia, and the Maltese hypogeum together constitute a cross-cultural record of what humans do in extended underground environments. Satyori's teaching on operating as other — on the internal reorientation that prolonged sensory deprivation induces — draws on the same comparative material and finds its archaeological ground partly in the Cappadocian record. Whether or not the shafts framed Polaris, the people who sheltered in these chambers for weeks at a time underwent one of the oldest documented forms of sustained sensory deprivation, and what that did to them is a question that connects the underground cities to the oldest contemplative traditions.

Further Reading