The dating of the La Tène period is very complicated, but is mostly
accomplished through diagnostic artefacts within a certain site and
extended across multiple sites and multiple regions (which introduces
important factors of time). Combinations of criteria are used --
including calibrated C14, but the latter is only of limited use. A good
example of this is to take a particular C14 date to see how useful it
might be in comparison to diagnostic artefacts of the same range.
Raftery, Pagan Celtic Ireland (1994, p.230) gives a sample of C 14 dates
for sites mentioned in the text. One of these is at the hilltop
enclosure of Dún Ailinne in Co. Kildare where an occupation phase 5, 6,
or 7 gave a sample (calibrated) of
390-30 B.C.. Even if it could be fully determined which phase this
sample came from, the range is far too great to date, say, a series of
brooches. Other dates of the site give an approximate habitation range
of 390 B.C. to 320 A.D. The artefacts have not been able to be dated
with any great accuracy -- a middle La Tène sword and a few Roman
period brooches (although one illustrated seems to be an identifiable La
Tène III type that might be focused somewhat better -- but as it appears
to be a British brooch, who can say when it got there). The rest of the
stuff is less dateable. 390 - 30 B.C. would span La Tène I to III
Fortunately, we have a few actual historical references from Italy and
other parts of the Mediterranean and elsewhere that can give us exact
dates. In Gaul, for example, there is Alesia. So what emerges from all
of this is a complex matrix of varying accuracy that is, unfortunately,
not too easily accessed by the general reader. Taking involutes alone,
Jope writes of these in the main body of the text in over three pages
and includes 27 footnotes. His illustrations and descriptions of such
adding more pages devoted to this sort of brooch. He also makes
comparisons and notes developments with details on other objects such as
the Newnham Croft armring and the Standlake scabbard. He mentions the
many different materials used in their construction including
red-painted sandstone (which must be later than than red coral it is
intended to imitate) an so on. From this, we can then take any of those
secondary objects, say the Newnham Croft armring, and its associated
finds in the same grave (pp. 45-7 and an additional 27 footnotes) and
track all of their influences and similarities, and so on and so on. It
follows that an effective critique of all of this is going to take up
even more room and then would follow the inevitable critiques of that
one until we might end up with something of encyclopaedic proportions
that no one is really going to understand -- all on involute brooches!
When Jope places the shallow profile involutes earlier on his time line
than the deeply profiled involutes, we can assume that he has good
reasons for doing so and that these include the presence of other
features that, on the whole, can be attributed as earlier or later
features by the comparison with other objects completely.
Following along from this, someone like Simon James comes along and says
"there was no unified Celtic culture" because he really believes that
to be the case, or John Collis comes along and says the same thing
because he really wants us to prove that he is wrong -- then the great
unwashed public starts to believe that "Celtic" is a naughty word. They
are then told to replace it with "Iron Age" (so as not to confuse the
dears) and then some of us Celticists who have achieved varying levels
of connoisseurship in our respective, and undoubtedly anal retentive,
specialties shake our heads and say "but...?"
What is really going on is just as scientific as C14 and
dendrochronology and the rest, but it is far more difficult to
demonstrate. I am sure that some who publish stuff about Early Celtic
Art could give even more precise dates than what appears in their papers
and books. It is just that, in order to build a tight case, it requires
an awful lot of writing. Better to just say 2nd century B.C. and leave
it at that -- especially if the innocent question that is asked will
steer us away from some of the things we really want to look into.
Things that we do not know yet.
Andy Holland wrote:
> Hi John,
> Thanks for your details and interesting comments.
> I'd not dream of questioning the relative sequence of the brooch chronology (I'm not qualified to do so as I don't know enough about them) but it's the tying of that relative sequence to an absolute date.....
> "The best that we can say about the involute type is that it is 2nd century B.C."
> Where did typologists get the 2nd Century BC date? You can't date the brooches themselves so it must have come from the dating of the archaeological contexts that the brooches were found in.
> NO problem with that - that's the way archaeology works. But what I want to bring a bit of caution to is..... the context may have been dated a long time ago - since when the dating technique used may have been shown to be flawed, inaccurate etc.
> The context date may get re-evaluated but often no one follows this through to change the date for the typology.
> Now for one date that's acceptable but many typology sequences have been built up over a very long period of academic study - so the original outline of the typology may have been based on very dodgy dating (perhaps even only ceramic typologies - which suffer from the same problem but on a much bigger scale!). This builds up the errors over and over to the point that the typology may really be a house of cards.
> There is really no way round this - except continued review of the typologies (which your doing) especially if the reviews occasionally re-examine the original dating technique to date each brooch type. It's an insane amount of work I know! Each dating technique for each brooch will have both an error range (re-determined in the light of current value of the dating technique) and a confidence - these can be combined in a standard mathematically and statistical method to form an overall confidence and error range for each section of a typology.
> Example (all hypothetical cases):
> Type b might be a very specific typological form found in only a small geographical area with really good multiple dating techniques that a baysian approach can really narrow the error and boost the confidence - those would therefore be very valuable in dating archaeological contexts.
> Type g might perhaps be a broad range of forms found in a wide area whose dating is based on really early C14 dating (before we began calibrating the dates) - so an uncal C14 date would give a very poor confidence and a large error range for this type making the presence of this type in an archaeological context of low value for dating purposes.
> Type j might be a form that was dated on the say of a learned professor of archaeology in 1946 based on his knowledge of pottery typologies of the region it was found in and the pottery typology was in that order because "he said so and he was the professor!" (we all know this happened and occasionally still does). The dates attributed to this type have no value at all - with no confidence in the date. This would be a good candidate for new research or a PhD thesis to re-evaluate the sequence around this type using more recent dating techniques.
> This is the sort of considered caution is what I meant by being careful with dating archaeology particularly in the British Iron Age.
> (the above is an example I use when teaching students about the difference between relative dating techniques (like artefact typology) and absolute techniques (like C14, OSL, TL U series etc)).
> Apologies if I'm teaching grannies to suck eggs!
> Not withstanding this your points about Ferrybridge are very thought provoking and I'll have to have a bit of a read of the report before I quiz my colleagues at Bradford.
> Take Care,
> Mr. A.D. Holland B.Sc. M.Sc. AIFA.
> Education Project Officer (11 - 18),
> Council for British Archaeology,
> Tel: 01904 671417
> Email: [log in to unmask]