1. Introduction
All historical disciplines deal with the fading traces of past interactions
between dynamic systems. Change, therefore, is omnipresent in historical
data, be it in the form of observable temporal dependencies between
elements in diachronic data sets or in the form of implicit dependencies
– hidden, so to say, in the temporal background of synchronous
data.
Dependency does not necessarily imply simple causation insofar as the
past state of such interacting systems does not usually fully determine
their future state but rather only restricts the possibilities for evolution.
Because of this lack of obvious cause-and-effect relationships between
historical events, the relevance of a specific instance of change in
a chain of interdependent developments is usually difficult to judge
even where data are abundant.
Conceptually, almost any ordered set of historical data, i.e. data belonging
to a single category, can be treated as a so-called Time Series, a statistical
term encompassing various types of chronological or – more generally
– serial data with dependencies. Although specific tools for quantisation
and statistical analysis of most types of historical data remain to
be developed, I believe that the concept of temporal dependency can
itself help greatly in the understanding of complex historical developments.
This essay aims at illustrating this point by re-examining the evolution
of Babylonian astrology in the 1st Millenium BCE in a broad sweep covering
various superficially unrelated types of data. Because of unforeseen
time constraints I had to give up my original research plan, which included
developing appropriate quantitative methods for each data type, and
can only present a rough, mostly qualitative sketch. Also, I lacked
the time to review the literature of the past two years and may therefore
be unaware of relevant new results. For an extensive recent bibliography
the reader is refered to Hunger and Pingree 1999.
2. Historical context
In the 2nd and early 1st Millenium BCE Babylonian Astrology was based
on the concept of dynamic divine intervention: celestial events like
eclipses or planetary conjunctions were considered interpretable signs
from the gods indicating that a more or less specific event would inevitably
take place on earth in the near future. Nevertheless, the gods left
some leeway for evading dire consequences. If, for instance, a celestial
event indicated that the king of Babylonia would become seriously ill
within a month, a substitute king could be appointed during the dangerous
interval that would then be stricken by illness instead of the true
king.
Both divine intervention and the possibility to react clearly stem from
a non-deterministic model of the cosmos.
As long as this type of astrology was in use, failure of a predicted
event to take place was blamed on the erroneous interpretation of the
observed sign. This is well attested by a lively academic discourse
on such matters in the late Neo-Assyrian period (Hunger 1992). While
a great reluctance to change the canonical system of omina is evident,
this nevertheless led to a fairly scientific approach to astrology insofar
as the canon was constantly tested against an expanding empirical database
and its exegesis changed rapidely. There is ample evidence for a systematic
and meticulous search for regular correlations between different types
of celestial and earthly events (Gerber 2000) and astrology is likely
to have been the driving force behind the establishment of an uninterrupted
series of astronomical observations extant in the so-called astronomical
diaries (Sachs and Hunger 1988, 1989). The expanding knowledge of the
periodic behaviour of many celestial bodies and events is at least in
the beginning most likely a by-product of this astrological research.
This early form of astrology is commonly called celestial divination
and was practiced routinely at the Neo-Assyrian courts. With very few
exceptions, texts attesting to its practice disappear towards the end
of the 7th century BCE. The following gap in attestations spans two
centuries, and the occasional reference to astrology during the time
of the Neo-Babylonian empire (626-539 BCE) testifies to a complete disinterest,
even contempt on the part of the court scientists (somewhere in Galter
1991…). Apparently, astrology was degraded to the status of folklore.
Only at the end of the 5th century does it quite unexpectedly reappear
in written sources in the entirely new form of horoscopy.
Horoscopes imply a completely different, deterministic model of physical
reality. If the celestial bodies visible at the time and place of birth
are considered to have a predictable influence on the life of an individual,
then no room is left for divine intervention later in life. The skies,
by this time, had become a clockwork rather than a writing board of
the gods. Parallely, astronomy had developed into an autonomous and
sophisticated predictive science.
It has repeatedly been suspected that these developments in the fields
of astronomy and astrology are somehow causally linked (Rocheberg 1998
et multa alia). To my knowledge, however, nothing beyond the obvious
has ever been said about the dynamics of this relationship; the ovious
being that the growing insight into the predictability of many astronomical
events at some point rendered obsolete the idea of celestial bodies
as signs.
3. Hens or eggs or neither?
The first known horoscopes appear in a period rich in scientific and
cultural innovations. After the incorporation of the Neo-Babylonian
empire into the Persian empire in 539, Mesopotamia ceased to be an independent
political entity and would, from then on, be ruled by foreign powers.
The fertilizing effect of the amalgamation of Babylonian, Levantine,
Persian and Hellenistic intellectual traditions renders the task of
differentiating between potential competing influences especially difficult.
Horoscopes are distinctly non-Babylonian if measured by 'traditional'
Babylonian (or, rather, Assyriological) standards, and would seem to
fit better into a hellenistic frame of mind. However, the first known
hellenistic horoscopes are chronologically younger, and there appears
to be a link - however vague - to the fairly straightforward evolution
of Babylonian predictive astronomy. Which of these or other possible
factors are the hens and which the eggs? Or, to rephrase the question
so as to fit the conceptual framework of this paper: Which are the relevant
dependencies between the different sets of available data, and what
are their characteristics?
4. Dependencies
4.1. Temporal distributions and dependencies
I restrict the following overview to four categories of data: Calendaric,
historiographic, astronomical and astrological. In the chart the –
very approximate! - frequencies of source texts vs time are indicated
in row A where this is relevant for the discussion. Row B represents
the approximate frequencies of contents vs time. The distinction is
important because contents extracted from earlier sources may be known
only from later compilations. In row C first attestations and relevant
developments are noted, and row D denotes major dependencies.
Astrology: the large gap in the sources has already been noted above.
Because the rare references to astrology during the Neo-Babylonian period
testify to a reluctant and somewhat sloppy use of traditional celestial
divination, we may savely assume that it had by then been discarded
by the courts' more serious academics, but also that horoscopy had not
yet been developed as an alternative. The gap therefore most likely
represents a true hiatus between two unrelated forms of astrology, and
not merely a random lack of sources.
Babylonian Historiography: under this heading I subsume all texts that
are traditionally refered to as the ‘Babylonian Chronicles’.
Their common denominator is a specific format: all are chronologically
ordered lists and a default entry consists of a date and a short description
of an event. Before ca. 610 no historically relevant relationship is
apparent between the events mentioned in a single text (Gerber 1998).
The statement that on day x of year y the king of Elam died may be followed
by a note that on day x + 1 of the same year a dog was seen peeing in
a temple and was chased away. After ca. 610 the Chronicles exhibit a
much clearer focus on related events. Most often these later texts can
be read as an historical narrative on events surrounding the royal family.
The selection of events to be included in the earlier Chronicles was
not based on the events themselves but on their date (Gerber 2000a).
The dates listed in these early texts show a highly significant, i.e.
a clearly non-random, correlation with the dates of astrologically significant
celestial events like planetary conjunctions. Because all known Chronicles
are late compilations from usually unknown older source texts (cf. Chart),
we may assume that the historical information in the early Chronicles
was extracted from astrological compendia that were used to search for
regularly occuring sequences of specific celestial and earthly events.
Although not directly attested in the known texts from the 7th century,
the well-documented academic discourse on celestial divination (Hunger
1992) proves the parallel existence of this type of research and the
canonical omina. In the later chronicles there are no non-random correlations
of this kind; they represent compilations from texts that had already
been written as historiographic documents and testify to an academic
interest in the past for its own sake already in the early 6th century.
Astronomy: astronomical observational texts, the so-called ‘Astronomical
Diaries’ (Sachs and Hunger 1988, 1989), are attested from the
8th century onwards with slight changes in focus but otherwise little
apparent variation. From 7th century-texts we know that these records
were kept daily by the court astrologers. The discovery of non-obvious
periodicities in certain celestial events like lunar eclipses was at
first most likely a simple by-product of the systematic search for astrologically
relevant regularities. The constantly growing observational data-base
allowed ever more precise numerical approximations and, at some point,
targeted searches for other periodicities that were no longer motivated
astrologically. Accordingly, during the 7th century, astronomical texts
concerned with prediction slowly become more frequent and testify to
the development of an autonomous branch of science that would thrive
and survive for centuries.
Calendarics: Contrary to the currently prevailing opinion the Babylonian
lunisolar calendar, which can be backwards-reconstructed with good precision
until 626 (Parker and Dubberstein 1956, for later corrections cf. the
references in Gerber 2000a), was fully periodic long before the introduction
of the metonic 19-year cycle in the early 5th century (Aaboe et al.
1991). This full periodicity was detected for the time after 600 by
means of a statistical technique called auto-correlation analysis (Gerber
2000b). The introduction of this new system of intercalation around
600 coincides with a 10 day shift in the mean beginning of the Babylonian
year (Gerber 2000a). For the time before 626 I proposed a systematic
reconstruction of ca. 60% of the Babylonian New Year’s Days that
suggests the use of a partially periodic intercalation system already
in the late 8th century. In a simulation study the attested calendar
dates before 539 were shown to be consistent with relatively simple
observational criteria for intercalation like, for instance, first visibilities
of specific stars or specific positions of the sun on the horizon (Gerber
et al 2002). Assuming such criteria were indeed used, the increasing
regularity in the calendar data during the 7th century and the shift
from partial to full periodicity would then correspond to the increasing
number of known periodic celestial events.
Lloyd Anderson, in the course of our correspondence on Babylonian calendars,
discovered a second distinct intercalation system that pushes the change
to full periodicity even further back in time to ca. 630 (see Appendix).
4.2. Determination of relevant influences
In the chart the dynamics of the discussed dependencies between the
four data sets are indicated by numbers refering to the individual sets:
‘1’ in row ‘3’ means ‘3 is dependent on
1’. For clarity’s sake I will call the data set corresponding
to ‘1’ in the example the ‘lead set’.
Within rows the sequence of numbers thus shows changes in the most relevant
influence excerted on a specific data set by another. These changes
can serve as chronological proxies for developments in the lead set
that cannot be dated in isolation.
Between rows the numbers within a horizontal slice show the hierarchy
of influence at a specific point in time: if ‘4 is dependent on
3’ and ‘3 is dependent on 1’ then ‘4 is dependent
on 1 by way of 3’. Here, ‘1’ is the lead set of the
highest order, i.e. the most influential set.
From this the following picture arises:
In the 8th and early 7th centuries celestial divination is the major
driving force influencing not only the development of astronomy towards
a predictive science but also, by way of its dependence on astronomy,
calendarics. From the perspective of the modern researcher celestial
divination even excerted a strong asynchronous influence on historiography
insofar as the format and structure of the hypothetical astrological
compendia were retained in the Chronicles that were later compiled from
them.
At some point the increase in known astronomically relevant periodicities,
and thus of the predictability of many celestial events, started conflicting
with the concept of dynamic divine intervention. This process inevitably
laid the groundwork for a radical switch to a deterministic world view.
In combination with a growing astrological database yielding no regular
correspondence between observed astronomical events and events on earth,
celestial divination and its associated world view were discarded entirely
within a very short time: around 630 the calendar system changes from
partial to full periodicity (see Appendix) and looses its indirect dependency
on astrology. At about the same time sources directly attesting to the
practice of traditional celestial divination disappear. The Babylonian
Chronicles continue to exhibit a strong correlation between mentioned
dates and astrologically relevant astronomical events until about 610,
and thus serve as a chronological proxy for the latest astrological
compendia. Only a few years later the source texts of the extant Chronicles
have changed to truly historiographic documents indicating that the
past was no longer viewed as a playground of the gods but rather as
a series of causally linked events that could be understood on their
own terms without recourse to divine intervention.
The change in the source texts available to the compilers of the Babylonian
Chronicles thus offers an extremely precise chronological marker for
a fundamental – and irreversible - cultural change in Mesopotamia.
Within this complex network of dependencies even the lag between cause
(insight into the predictability of the skies) and effect (shift to
a deterministic world view) can be determined as being smaller or equal
to 20 years.
In this new cultural context established at the end of the 7th century,
historiography and – much later, with the introduction of the
entirely schematic Metonic cycle – calendarics regain their independence
from astronomy. The vacuum created by the disappearance of traditional
celestial divination was eventually filled by a new form of astrology
that was entirely compatible with the new world view: horoscopy. 5.
Conclusion
Despite the shortcomings of mostly qualitative background of this study,
I hope to have shown that statistics, especially Time Series analysis,
or, to a lesser degree, even bare statistical concepts are valid tools
for structuring and analysing complex historical developments as well
as for finding and dating instances of cultural change. In the case
treated here, namely the interplay between Babylonian astrology and
astronomy, they may even be the only tools leading to controlable results
that allow one to go beyond learned guessing.
References
Aaboe, A., Britton, J.P., Henderson, J.A., Neugebauer, O. and Sachs,
A. 1991. Saros Cycle Dates and Related Babylonian Astrionomical Texts
(TAPS 81.6). Philadelphia.
Galter, H.D. (ed.) 1991, Die Rolle der Astronomie in den Kulturen Mesopotamiens.
Beiträge zum 3. Grazer Morgenländischen Symposion (23-27.
September 1991). Graz.
Gerber, M. 1998a. Die Inschrift H(arran)1.A/B und die Neubabylonische
Chronologie, ZA 88: 72-93.
Gerber, M. 2000a, A Common Source for the Late Babylonian Chronicles
Dealing with the 8th and 7th Centuries, JAOS 120.4: 553-569.
Gerber, M. 2000b, The Babylonian Civil Calendar 731-626 B.C. Evidence
for Pre-'Metonic' Periodic Intercalation Patterns, in: Esteban,C. and
Belmonte, J.A. (eds.), Astronomy and Cultural Diversity (Proc. Oxford
VI & SEAC 99). Santa Cruz: 244-248.
Gerber, M., Grasshoff, G. and Donatowicz, J. 2002, Observational Criteria
for Intercalation prior to 539 BCE, Research Report, Berne
(http://www.ane.unibe.ch/projects/uos2001.pdf).
Hunger, H. 1992, Astrological reports to Assyrian kings (SAA 8). Helsinki.
Hunger, H. and Pingree, D. 1999, Astral Sciences in Mesopotamia (HdO
I/44). Leiden.
Sachs A., Hunger, H. 1988, 1989, Astronomical Diaries and Related Texts
from Babylonia. I, Diaries from 625 B.C. to 262 B.C; II, Diaries from
261 B.C. to 165 B.C. Vienna.
Parker, R.A. and Dubberstein, W.H. 1956, Babylonian Chronology 626 B.C.
– 75 A.D. (Brown University Studies 19). Providence.
Rocheberg, F. 1998, Babylonian Horoscopes. Transactions of the American
Philosophical Society 88.1.
Appendix by Lloyd Anderson (to be shown at WAC5, showing change from
one 19-year lunar system to another 19-year lunar system)
……
|