The Great Sphinx of Giza

The Sphinx lies on the eastern edge of the Mokattam Formation outcrop that forms the Giza plateau, about 500 metres east-south-east of Khafre's pyramid. Gauri's stratigraphy divides the exposed rock into Member I, a hard reef limestone forming the base and the ditch floor; Member II, seven alternating softer and harder beds from which most of the body is cut; and Member III, harder at the top, from which the neck and head are carved and which was quarried away across the plateau for building stone⁠[12]⁠[5]. The ditch floor is at about 20 metres above sea level, near the range of the historical Nile flood⁠[6].

The statue has been buried in sand for most of its history and cleared at least five times: by Thutmose IV about 1400 BC, who set up the granite Dream Stele between the paws and whose name is on mudbrick walls around the enclosure; in the Saite and Roman periods; by Caviglia in 1817 to 1818, who found the stele and the beard fragments; by Mariette and Maspero in the nineteenth century; by Baraize from 1925 to 1936, who exposed the Sphinx Temple and added much of the present masonry; and by Selim Hassan in 1936 to 1938⁠[1]⁠[13]⁠[2]. The body is clad in limestone blocks of several phases, the oldest of which Lehner attributes to Thutmose IV; the original carved surface survives beneath them⁠[1].

The ARCE Sphinx Project (1979 to 1983, field director Mark Lehner) produced the first complete record: photogrammetric elevations and a plan at 1:50 by Ulrich Kapp, maps of the two temples at 1:100, geological and topographic maps of the enclosure at 1:200, sections of the masonry at 1:10 and 1:20, about 7,700 photographs and a water-table profile for 1981 to 1982. The archive was published through Open Context in 2017 under a CC BY 4.0 licence⁠[14]⁠[15]. Lehner's 1991 Yale dissertation is the full written publication of the survey⁠[16]. Hawass's team laser-scanned the statue around 2010 for conservation; those data are not public⁠[17]. The Giza Project at Harvard publishes a 3D model of the Sphinx for its reconstructions, without stated accuracy or download terms⁠[18].

No Old Kingdom text names the Sphinx; texts about it begin in the New Kingdom⁠[1]. The best known is the Dream Stele, a 3.6-metre granite slab reused from a lintel of Khafre's mortuary temple, set up in Thutmose IV's first year (about 1401 BC). It tells how the prince slept in the Sphinx's shadow, was promised the kingship by the god Horemakhet if he cleared the sand, and did so. A damaged line contains the syllable "Khaf", read by some as Khafre's name, but the passage is fragmentary⁠[13]. The Inventory Stela (Cairo JE 2091), found by Mariette in 1858 in the Isis temple beside the Great Pyramid, says Khufu found the temple of Isis already standing beside the "house of Harmakhis", the Sphinx, and rebuilt it. It is a Saite (Twenty-sixth Dynasty) composition, written about two thousand years after Khufu, which names him by his Horus name, gives Isis a title she never held and makes Henutsen a king's daughter; Egyptologists treat it as a claim of antiquity by the Isis temple's priests⁠[19]⁠[20].

The beard is the Sphinx's one detached part with a museum history. Caviglia recovered fragments of a plaited limestone beard in 1817 to 1818; one piece (78.7 cm high) is in the British Museum as EA58, and other pieces are in Cairo. The style is New Kingdom, and the pieces are read as an addition of Thutmose IV's restoration rather than part of the original carving, though Digital Giza files the fragments under Dynasty 4⁠[21]⁠[22]. Fragments of the Sphinx's uraeus, the royal cobra on the brow, were found with the beard⁠[2]. A life-size statue of Khafre, found by Mariette in the ruins of the Valley Temple, is the basis for reading the Sphinx as part of his statue programme⁠[2]⁠[1]. In 1978 Hawass and Lehner found three stone blocks abandoned in a corner of the Sphinx quarry while being dragged to the Sphinx Temple, which they read as a sign that the workers stopped abruptly⁠[2]. A passage under the Sphinx, opened by Baraize and re-examined by Hawass and Lehner in 1980, proved to be a shallow cut with no chamber⁠[23].

Three lines of evidence tie the Sphinx to Khafre. The Sphinx Temple's core blocks carry the same fossil beds as the Sphinx ditch, so the temple was built from the ditch's quarry⁠[1]⁠[2]. The temple's plan, core-and-granite construction and alignment match Khafre's Valley Temple beside it. A drainage channel running along the north side of Khafre's causeway opens into the upper corner of the ditch; Selim Hassan reasoned that the builders would not have let it drain into an existing ditch, so the ditch was quarried after the causeway, late in Khafre's reign. Schoch countered that the rear of the ditch was cut much later than the front and sides, so the channel dates only the west wall; Lehner replies that Schoch then cannot also cite the west wall's weathering as evidence of age⁠[7]. The Sphinx stands at the end of that causeway as the valley element of the pyramid complex, and the Khafre statues from the Valley Temple give the only royal face for comparison⁠[1].

The geology is also well recorded. Gauri's group published the bed-by-bed porosity and durability of the limestone and showed that the softer beds of Member II recede faster whatever the agent; they attributed the modern flaking to salt crystallisation driven by moisture and daily temperature cycles⁠[5]⁠[24]. The deep rounding of the ditch walls is agreed by all sides to be real; what is disputed is the agent and the time it took. Harrell measured the enclosure floor against flood levels and the water-holding capacity of the sand that filled it for most of 4,500 years, and argued that wet-sand burial and repeated flooding produce the observed decay without any pre-Old Kingdom time⁠[6].

Schoch's seismic profiles found a weathered low-velocity layer beneath the ditch floor that was deeper on the sides and front than behind the rump, which he read as longer exposure of the sides; Harrell pointed out that the velocities behind the Sphinx were lower than at the sides, the opposite of what that reading needs, and that the layer may be a depositional reef facies⁠[3]⁠[6]. Reader's 2001 Archaeometry paper accepted runoff erosion in the ditch and dated it by the quarrying history of the plateau above, which drew a published comment and reply⁠[8]⁠[25]. The 2023 NYU experiment is a laboratory analogue only: 10-centimetre clay mounds in a water tunnel, at Reynolds numbers far below natural wind-blown sand, and the authors say their results "hardly resolve" the Sphinx⁠[10].

Khafre, about 2500 BC. The Sphinx was designed as part of Khafre's complex and carved with the Sphinx Temple in a single quarrying operation; the head is his portrait; the weathering is the product of soft beds, salt, wet-sand burial and flooding over 4,500 years, and tells nothing about age beyond that⁠[1]⁠[5]⁠[6]. This reading predicts that no feature of the Sphinx or its enclosure can be shown to pre-date Khufu's quarrying, and that the Sphinx Temple's masonry and the ditch are contemporaneous. Lehner argues both hold; no pre-Khufu material has been published from the enclosure⁠[7].

A much older core, 5000 BC or earlier. Schoch holds that the rounded profile and deep fissures are rain erosion from the wet Sahara of the early Holocene or the end of the last Ice Age, that the seismic weathering depth implies 7,000 to 12,000 years of exposure, that the head was recut in dynastic times (which is why it is small for the body), and that a lion or lioness was the original form⁠[4]⁠[3]. It predicts datable pre-dynastic human activity in the enclosure, erosion features in Member I and II that no Old Kingdom process can produce, and a seismic weathering gradient consistent across the ditch. None of these has been shown; the seismic gradient is contested, and no pre-dynastic material has been found in the enclosure⁠[6]⁠[7].

Early dynastic, about 2700 to 3000 BC. Reader accepts runoff as the agent but places it in the Early Dynastic Period, when the plateau's drainage ran into the Sphinx area before Khufu's quarries intercepted it; he argues Khafre's Valley Temple is later than the Sphinx Temple, that Second Dynasty masonry skills were sufficient, and in 2026 that the body was carved for the lion goddess Mafdet and the head recut by Khufu⁠[8]⁠[9]. It predicts that the ditch's runoff features pre-date Khufu's quarry cuts stratigraphically and that the Sphinx Temple is earlier than the Valley Temple. Egyptologists have not accepted the stratigraphic argument. The book had a favourable notice in Ancient Egypt magazine in April 2026, and Peter Lacovara's review is in press in the American Journal of Archaeology⁠[26]⁠[27].

A natural yardang shaped by the builders. The NYU experiment shows that a hard inclusion in an eroding mound yields head, neck, back and paws under a steady flow; the reading is that a wind-carved outcrop suggested the lion and was finished by hand⁠[10]. It predicts a pre-existing landform in the plateau's geomorphology at the Sphinx's position; the quarry evidence is that the body was freed by cutting a ditch into a flat bench, which leaves little room for a free-standing yardang⁠[1].

Little erosion at all. Schneiker's 2025 poster proposes that the rough body was shaped by pounding during construction and faced with a veneer, that sand protected it, and that looting of the veneer exposed the rough core now read as weathering⁠[11]. It predicts that the "weathered" surfaces carry tool marks under the masonry and that the fissures are quarry features. Lehner's record of the core beneath the earliest cladding is the test; no paper behind the poster had appeared by October 2026.

What it represented. The head is small for the body, which every side notes. The usual reading is that the sculptors reserved a block of good Member III stone for the head and carved the body to the rock available, or that the body was lengthened to pass a fissure⁠[1]. Robert Temple argued in 2009 that the original was a jackal, Anubis, with the present head recut under the Middle Kingdom; Schoch and Reader argue for a lion or lioness head recut in the Old Kingdom⁠[28]⁠[4]⁠[9]. Each predicts a different original neck line and proportions beneath the New Kingdom and later cladding, and none has been tested against the ARCE elevations.

The age question turns on stratigraphy that already exists in map form. The ARCE archive's 1:50 elevations, 1:200 geological maps and masonry sections allow the Member II bed recession to be measured bed by bed around the whole body and ditch and compared with the same beds in Khufu's and Khafre's quarries; if the agent was early Holocene rain, the pattern should not correlate with burial depth and flood level, and if it was wet sand and salt, it should⁠[15]⁠[6]. A repeat seismic survey of the ditch floor with modern equipment, published with the raw traces, would settle the weathering-depth argument that has stood unresolved since 1992⁠[3]. Luminescence or cosmogenic-nuclide dating of the quarried ditch faces, where unweathered rock was first exposed, would give an exposure age independent of both camps.

The head question needs the data under the cladding: a published 3D model of the core body from the ARCE photogrammetry or the 2010 laser scans, with the neck and chest profile analysed for recutting. The Khafre attribution would be strengthened by a Fourth Dynasty text naming the statue, which has never been found, and would be overturned only by securely dated pre-Khufu material in the enclosure or under the Sphinx Temple.