Predicting the Future: The Delphi Tradition


Some Background to Delphic Polls

<-- The Tholos (rotunda) in the Athena Pronaia temple complex at Marmaria, the Gateway to Delphi.
In strictness, the modern "Delphi Technique" does not set out to attempt to forecast the future, although it has been used to arrive at such a forecast, but rather it is "an interactive and personality-free team approach to decision making" (McNamee). It was verified and developed in its present form by the Rand Corporation (1964) for both normative (target orientated) and exploratory long range forecasting.

A Delphic Poll makes decisions about a scenario by testing the opinions of a panel of members. It is characterised by three features which distinguish it from other consensus-achieving group forecasts:

This requirement for a "Statistical Response" and the subjective, fuzzy nature of the data have given rise to some questions about the pictorial method chosen to display the results of the 1978 questionnaire (see the original articles). "Statistical Layout" seems somewhat too formal a description for the results presentation, since I'm not at all sure that the pictures have any real statistical validity - it just seemed to be a reasonable way of "blurring" the answers - that is of presenting them in a way which didn't give a spurious feeling of great accuracy. The method was based on one used by the late Chris Evans (NPL, author of "The Mighty Micro" etc. etc.) and since he was a psychologist I assume there must have been some theoretical basis.

Briefly, if one had this distribution of responses ("votes"):-

2 8 15 4 8 2 3 20 1 1 1 0 6 1 1

then ignore the outliers, calculate a simple weighted mean (~1989) and draw a polygon (triangle on thin rectangle base) with standardised height at apex and at base to cover the spread (apex is at accurate mean) thus:-

These "barn-shaped" cross sections give some immediate visual indication of the degree of consensus of the responses - if the shape is squashed up, it shows that people are generally in agreement (though look at the size of the "never" vote. Tight distributions, however, don't mean that the opinions are "right", just that people are in agreement about them). Skewed shapes show some polarisation of opinion and so on. Perhaps I should have explained this better in the recent "Spectroscopy Europe" article, but there was limited space and I didn't want to get bogged down in statistical arguments!


The Delphic Oracle

The Oracle of Apollo at Delphi - the most famous in Greece - held a remarkably long currency. Its earliest beginings lie in the Dorian Invasion and the Herakleidai, descendants of a Mycenaean dynasty. This link back to Herakles (Hercules) is typical of the entanglement of the real with the mythical - it was reported that Hercules himself did not become immortal until the Oracle said so (Flavius Arrianus, "The Life of Alexander the Great") In its situation on the slopes of Mount Parnassos, in the angle of the Phaedriades ("Shining Rocks), the sacred precinct enjoys an impressive setting - "a theatre of forbidding precipices, with springs, crevasses and exhalations" as befits the Omphalos - the centre of the world, where the eagles of Zeus met.

At the height of its power in the 6th and 7th centuries BC, the temple of Apollo which housed the oracle was richly endowed and enjoyed great international fame (in contrast with other sanctuaries such as Olympus, which were largely nationalistic in character). International benefactors however, of whom perhaps the most famous was Croesus, last king of Lydia (560-546 BC), enjoyed no greater protection from the oracle's ambiguity than others. In his battle with the Persians at Sardis, a great empire was destroyed, as the dutifully consulted Oracle foretold (thus encouraging Croesus to fight), but it was Croesus' own empire that was lost to Cyrus' advancing forces.

Nevertheless the Oracle became ever more rich and influential, and the Temenos contained magnificent treasuries from Corinth, Sikyon, Knidos, Syracuse, Athens, Siphnos, Kyrene and many places further afield. As might be expected, such riches attracted envy and controversy and in their wake, wars. A series of "Sacred Wars" over the next two hundred and fifty years, each resulting in changes of control of the city of Delphi and its most famous adjuncts, began the Oracle's long decline. Famous "power brokers" took an interest from time to time - Xerxes tried to loot the temple, Philip of Macedon (the father of Alexander) took over for a while, and attacks by Gauls, repulsed in a similar supernatural manner to the Persians, were finally replaced by the stability (and disinterest) of the Romans.

The Oracle itself still held some sway "world wide", despite its strong Doric prejudices - Nero seized 500 bronze statues from the temple in a fit of rage when the Oracle condemned his matricide. Domitian somewhat restored it - the visiting Pliny counted more than 3000 statues. Constantine carried off the sacred bronze Tripod of Plataea to decorate the Hippodrome in his new capital (there are those who measuring the feet of the bronze horses in St Marks in Venice and finding a fit with the "hoof marks" of the missing horses which drew the Chariot of Helios, would say that Constantine carried them off too - for the present-day horses certainly reached Venice from Constaninople). Julian consulted the Oracle: Theodosius finally abolished it in about 385 AD and the site lay ignored until the 17th century. A visit by a "disappointed" Byron and the excavators of the 19th century began its re-emergence as a modern tourist centre.

In its hey-day, enquirers who sought to "work the Oracle" had first to travel there. On arrival an animal sacrifice (sheep, goats, and boars were common offerings) would indicate whether the omens were favourable, and whether to proceed to the temple where they would wait their turn (decided by casting lots, unless a previous "Promanteia" - prior right of consultation - had been received (presumably at a price!) from the Delphians) A question, written on lead tablets (many of which have been recovered) was handed to an official who would consult in turn with the Pythia, or priestess, who delivered the Oracle. After ritual purification in the Castalian Spring (in the ravine betwen the temple and Marmaris) drinking the waters of Kassotis and chewing laurel leaf, the priestess took her seat on the tripod above the (seismic) chasm in the inner temple. There, breathing the volcanic exhalations, she would utter incoherently, and these utterances would be translated into hexameter verse by the poet in waiting for delivery to the enquirer. The interpretations were always obscure and frequently ambiguous, and the enquirer often returned more mystified than he came.
Despite this, and the famous cases of those apperently misled by the pronouncements, at least one philosopher (Strabo) is on record as saying that "of all the oracles of the world, (Delphi) had the reputation of being the most truthful".

The drawing of parallels between the functioning of the Delphic Oracle and those of the Internet are left at this stage to the reader.

A portion of the frieze from the Siphnic Treasury
The battle between the Giants and the Gods

[home] This document created and maintained by Acolyte Science
For comments, suggestions, or contributions,
E-mail to
© Acolyte Science, 1995
Last Update, 10-Nov-95