Macedonian Civilization

Theodore G. Antikas
History Department, University of Washington, Washington State, USA
Summary. Five horses and two dogs buried in a human cemetery dating to the 4th-7th c. BCE were unearthed in 2003-5 during a rescue archaeology project at Sindos, southwest of Thessalonike, Macedonia. The good state of preservation of both human and animal skeletons due to the sandy alluvial deposits of the Rivers Axios and Haliakmon that cross the area was rare enough to prompt a detailed archaeozoological analysis. In addition to routine macroscopic
and microscopic examination to determine sex, age, height, use and palaeopatholgy, carbon dating and multiple stable isotope analyses were performed by two independent laboratories: (a) the “Democritus” Nuclear Research Ctr.
Archaeometry Unit, Athens; (b) the Carbon Data, FLA Both laboratories came up with identical data concerning stable isotope analyses which indicate that a portion of the horses’ food was fish-related. The unexpected find is discussed in view of Herodotus’s Histories in which he recorded the habit of ancient Macedonians to feed their pack horses with fish. Although more extensive sampling and research is needed to prove or disprove Herodotus, this find indicates that historians who referred to him as “father of lies” may need to reconsider their views.
Keywords: horse, stable isotopes,Macedonia (not Greece).

Introduction. “The houses of the <Macedonians>
lake-dwellers are actually in the water … on platforms
supported on...piles and approached from the land by a
single narrow bridge… Each member of the tribe has his
own hut…with a trap- door opening on to the water…
Their horses and other pack animals they feed on fish,
which are so abundant… that, when they open the trapdoor
and let down an empty bucket on a rope, they have
only a minute to wait before they pull it up again, full.
The fish are of two kinds…” (Herodotus, Historiae 5.16).

The burial of nobles, heroes and Olympic victors with
(their) horses was first documented in Homer’s Iliad. That
Mycenaean practice was to last for a long period of time
as reported by later writers such as Herodotus and many
others. Hundreds of such burials dating from the 16th to
the 3rd century BCE have been found in tombs spreading
from W. Thrace to the Peloponnese, in the last thirty years
(Figs. 1-5), thus confirming the reports of ancient writers
almost verbatim. On the other hand, the possibility of
equines being carnivorous (or at least omnivorous), some
time in their fifty-million-year evolution, was thought to
be a myth and was discussed only scantily in the works of
ancient writers, among which Eurypides and Pausanias.
The latter do mention “man-eating” horses and mares,
turn crazy until they are finally “tamed” (or
domesticated?) by legendary heroes. Herodotus goes as
far as to support the idea of fish eating horses as seen on
the caption of this introductory note, and his phrase has
caused skepticism to more than a few modern scholars,
who claim that he made “gross overstatements” and add
these in their armory aimed at discrediting him.
The existence of man-eating or fish eating horses in
archaic times, whether it be true or false, is a legitimate
hypothesis that could only be tested in a scientific
multidisciplinary approach by archaeologists,
archaeozoologists, geneticists, biologists and nuclear
physicists. In this context, the Aristotelian University of
Thessalonike; the 17th 16th and 4th, Ephorates of Classical
Antiquities at Pella, Thessalonike and Nauplion,
respectively); and the Nuclear Research Center
“Democritos” in Athens adopted a five-year project
proposed by the author, in 2003. The first extensive
skeletal analyses for stable isotopes (oxygen, carbon,
nitrogen) from bones and teeth of horses buried at Sindos
(Fig. 6) took place in the period of 2003-2005. The results
tend to confirm the existence of protein-eating horses in
the past, just as Herodotus and other ancient authors
mentioned above had claimed. The evidence obtained
from the Sindos horse skeletons is the subject of the
present paper.

Materials and Methods
1. The Sindos Horses
The 47 humans, five horses and two dogs unearthed at
Sindos date from the 7th to the 3rd century BCE. As the
MNI of adult men was estimated to be 17, the high ratio
of horses per man (1:3) suggests that Sindos may have
been a city where cavalry men (asthetairoi) were
recruited for the necessities of wars during the mid-4th
century BCE (Philip’s era). This because the maximum
ratio found in other necropoleis in Macedonia was 1:200.
In addition, Sindos lies between the rivers of Axios and
Gallikos in Central Macedonia, near the lake area of
Volvi (Fig. 6). In other words it could be the rich area
Herodotus mentions in his passage in reference to King
Darius’ march through the plains of Macedonia in
the turn of the 5th century BCE. Horse bones and teeth
were handled with gloves, cleaned with distilled water,
photographed, X-rayed, packed in acid free and/or
aluminum paper, and those in perfect condition were
selected for radiocarbon and stable isotope analyses
following standard techniques.

2. Carbon and nitrogen stable isotopes
Carbon/nitrogen isotopes and diet
Carbon and nitrogen isotopic values from human
and/or animal bone collagen are thought to reflect the
diet, particularly that of dietary protein. Using such data,
it is often possible to draw conclusions about:
• The levels of animal protein (meat and dairy
products) in the diet, as compared to terrestrial plant
consumption, i.e., the identification of omnivorous/
carnivorous dietary behavior.
• The consumption of plants following the C4
photosynthetic pathway (or protein from animals
consuming such plants) compared to the consumption of
C3 plants.
• The consumption of marine and freshwater dietary
resources, in particular fish.
Values are averages of two replicates in separate
batches using a “Europa Scientific Geo 20/20” isotope
ratio mass spectrometer coupled to a “Roboprep”
elemental analyzer. Controls analyzed were calibrated to
international standards. The analytical precision is
considered to be ± 0.2‰ for both δ13C and δ15N. The
collagen extraction employed ultra-filters, so that the
collagen analyzed is that with a higher molecular weight.
This is expected to remove contaminants and degraded
collagen, but the process considerably reduces the yields
for the samples. The quality indicators given in the table
rely more on the C:N ratio and the element percentages
than on the yield percentages. On this basis, the collagen
is considered to be of appropriate quality to provide
acceptable data.
Two independent laboratories were chosen to test the
horse samples in a double blind protocol in order to avoid
unnecessary sampling and preserve the anonymity of each
sample and/or laboratory. Identical batches of bone and
teeth samples were shipped to (a) the Stable Isotopes
Laboratory at the University of Bradford and (b) the
Archaeometry Lab at the Nuclear Research Center,
Athens. Finally, the bone and teeth samples sent to both
labs were given code numbers, i.e., Sin-1, 3, 4, 5 (horses),
Sin-2 (dog), Sin-6 (cow).

Results
I. Archaeozoological Analyses
Horse # 1
Stallion aged over 18 years and in poor conservation,
probably due to the fact that it is the earliest burial dating
to the 7th-6th century BCE. Carbon dating supported the
initial, indirect dating based on a bronze bridle artifact
probably made locally or at neighboring Pella. Under the
horse’s 4th lumbar vertebra (Fig. 7) small fragments of
mandibles and teeth of capra/ovis, and of sus scrofa were
found indicating that they were probably “leftovers” from
a sacrifice or a funeral dinner, as there were signs of pyres
around the tomb. As this stallion was the oldest among
the five horses found at Sindos, and because carbon
dating places him in the 7th-6th century, samples were
taken for stable isotope analysis.

Horse # 2
Adult mare aged over 15 years and belonging to a
microsomatic type of autochthonous Macedonian horses, such
as the Skyros pony Her estimated height does not exceed
118 cm (12 hands). Pathological finds at the interphalangeal
and tarsal joints (Figs 8-9) strongly indicate
the use of this mare as a cart animal, which makes her not
a very good candidate for stable isotope analysis.

Horse #4
Aged 3-3.5 years, this was the youngest horse found at
Sindos but its sex could not be determined. The age was
calculated by the incomplete fusion of the epiphysis to the
head of the humerus and/or the olecranon, processes
(completed at the age of 4-41/2 years-Figs. 10-11). Other
than the surprising young age of Horse #4, a very
interesting detail concerning its burial site is the proximity
to a young man’s skeleton. As shown by arrows
indicating the two graves on Fig. 1, the distance of ca. 55
cm, and the position of the horse across the dead man
The man of T-34 was aged 20-25, and was buried with
a wreath on his head, a strigil, a silver coin and a silver
vase. Such finds are usually funeral offerings to cavalry
officers and/or victors in equestrian events and have been
documented by ancient historians such as Herodotus.
Speaking of the burial of the Athenian Kimon, a triple
victor in the tethrippon races at Olympia (536, 532, 528
BCE), Herodotus uses the term ‘enanti’ (Gr: across) to
describe the burial position of Kimon’s mares in his tomb,
just as in the case of T-34 and Horse #4 at Sindos. Hence
it is logical to assume that T-34 may have been a cavalry
officer and/or a successful anabates (rider). Finally,
seashells accumulated to construct a surrounding wall
may indicate a yet unknown cult practice, or simply a
funeral dinner consisting of seafood, since Sindos is
proximal to the delta of the Axios as it pours into the
Thermaic Bay. The abundance of sea-derived food around
this young horse and the postulate of its probable
ownership by an affluent hetairos (companion) of the
Macedonian cavalry were the reasons for sampling its
bones for stable isotope analysis. Surprisingly, another
find in a Macedonian tomb at Hagios Athanasios (a
neighboring site to Sindos) where a rich anabates (rider)
bearing a wreath and buried with horse tack across him
bears a striking similarity to the T-34 and Horse-4
common burial at Sindos as shown on Fig 12.

Horse #5 with Dog #2
Adult mare aged 16±1 and measuring 140±2.5 cm at
the withers. Her somatometrics indicate a phenotype of
the native Thessalian horse breed to which Bucephalas
belonged (Fig. 13). Pathological signs on her tarsal joints
suggest a chariot or a polemic (war) horse. This
hypothesis is substantiated by her size and the excellent
state of preservation. These two parameters made her a
candidate for stable isotope sampling. Seashells and burnt
coccygeal vertebrae found proximally to her legs and in
pyres next to the dog, strongly suggest a rich funeral
dinner. Dog #2 buried close to her hind legs on the northsouth
axis was a young bitch aged 18-20 months, as
indicate the dental attrition levels and the non-fusion of
humeral heads. Her height was 30±2 cm, indicating a
small type of hunting dog referred to as “Ghekas”, whose
provenance is north-northwestern Macedonia. Hunting
dogs were often depicted on Macedonian coins (Fig 14), vases
and frescoes. A most impressive scene is found on the
fresco adorning the antechamber of the Royal Tomb II at
Aigai where no less than nine dogs are shown hunting
with Philip, Alexander and their companions (Fig 15).

II. Radio-isotopic Analyses
Carbon-14 Dating
Bone and tooth samples of four horses coded for
laboratory purposes as SIN-1, SIN-3, SIN-4 and SIN-5.
Samples from Dog-2 and a cow were added for
comparative purposes and coded as SIN-2 and SIN-6,
respectively. Carbon dating placed Horse-1 to the 6th-7th
c. BCE, and Horses 3, 4 and 5 to the 5th, 4th and 3rd
centuries BCE, respectively.
Stable Isotopes
Table I lists the results and Figure 19 depicts the same
results in chart form.
The δ13C, δ15N values coupled with the carbonnitrogen
ratios indicate that at least 18-22% of the horses’
feed may have consisted of protein of fish origin.
However more analyses are necessary to compare the
equine data with those of herbivorous and/or omnivorous
animals (bovines, capra/ovis, canines, felines) from the
same time period and the same region, i.e., the Central
Macedonian plains.
Some δ18Ο values have also been obtained from both
ancient and modern equines so as to establish water
precipitation in the Sindos plain as opposed to
mountainous areas in Central Macedonia some 2500 years
ago. However the results are outside the scope of this
paper.
Discussion and Conclusions
As Hesiod and Homer, Herodotus has also been
criticized for including ‘mythical’ and ‘exaggerated’
realities in his Histories. Recent multidisciplinary studies
by geneticists, archaeologists, archaeozoologists and
physicists find elements of truth in more than a few of
these exaggerations. As surprising Herodotus’ statement
on fish-eating horses may seem a less skeptical approach
would lead scholars to consider it from a more scientific
angle. Firstly, one has to wonder why at least three
ancient sources mention man-eating or fish-eating
equines. Secondly, the question as to why E. caballus a
perissodaktylon (odd-fingered) like man has four sizable
canine teeth if it where purely herbivorous, needs an
answer. Last but not least, one has to consider whether
Herodotus, an invitee at the Macedonian court, should
refer to “fish-eating horses and other pack animals”
unless he witnessed or heard from locals that this was a
fact. In another instance, Herodotus wondered why the
lions at the River Axios “ignored other living creatures
and set only upon the camels-beasts never seen…
before”. Scholars have discredited the ancient historian
by claiming there were no lions known to have existed in
Macedonia. However, recent archaeological digs in
Central Macedonia have unearthed skeletal elements of
both lions and camels thus confirming both Herodotus
and Xenophon, who went as far as to describe how lions
were hunted at his time.

The archaeozoological and stable isotope data re the
Sindos horses tend to confirm there is truth in Herodotus’
claims. A next step in the pursuit of truth would be to
expand our study so as to include equines found in
different regions of Macedonia (Thrace, Thessaly, Mycenae)
dating to earlier periods, i.e., Minoan, Mycenaean, Iron
and Classical Age. In addition, there is need for more
control data obtained from purely herbivorous and
omnivorous animals, so as the stable isotope values
become comparable. Last but not least, an experimental
protocol should test modern horses’, mules’ and donkeys’
capacity to accept or reject eating animal protein feed(s)
of meat or fish provenance. Unfortunately, recent
developments in regard with BSE have impeded this
phase of experimental research although our preliminary
data on the palatability and acceptance of fish meat by
horses supports the stable isotope finds.
In conclusion, Herodotus does not seem to have
recorded myths or hearsays in speaking of “fish-eating
equines” or “camel-eating lions” in the rich plains of
Central Macedonia of the early 5th century BCE. Just as
Homer before him was not referring to a mythical Troy
but to a real citadel, Herodotus was seemingly telling a
story in his capacity as the father of history, but certainly
not as the “father of lies”.?

http://www.lva.lt/vetzoo/data/vols/2008/44/pdf/antikas.pdf