Interesting thread. Based upon all of the discussion, it seems clear to me that the only things that are truly historically accurate are the actual antiques themselves.

When it comes to modern reproductions, historical accuracy is defined by the individual collector, presumably after 'enough' research (also defined by the individual collector) to make his or her own conclusions. What may satisfy one person as “historically accurate” may be quite unacceptable to another. The more 'extreme' collector’s requirements may be considered by others as hysterically accurate :lol: .

It is all in the individual perspective (and budget), and these both are likely to change over time. I know my standards have changed in the last few years.

This topic has been promoted into a Spotlight Topic.

100% historical accuracy is only possible to approach " 99.99 % " if our information is complete and reliable: There is always that little gap in knowledge that one must fill with the most probable.

How much of this accuracy one wants or insists upon may vary from person to person and in a case by case basis.

A lot of history is guesswork or reconciling conflicting primary sources or differences of opinion in the interpretation of sources.

Can anyone say that when reading yesterdays News we would all agree on what happened 100%. :lol:

Well, an original object could be said to be the ultimate " accurate " object ! But what if this " pure " object has been modified over the centuries i.e. 11th Century blade remounted in the 15th Century ? Or simply changed by centuries of wear or repair !

Well, I think you all get the point that I tend to be wary of absolutes ! This doesn't mean that one shouldn't try to be as accurate as possible as possible with modern replicas if that is one's interest and goal.

History isn't History: History is our best accepted and current version of it.

Hmmm.., maybe I should explain a bit.
I am a keen horse rider and have been interested in applied archeology for 35 years. A close friend of ours is right now coming over from Belgium for weekend of rest and brainstorming: she is a professional 'commercial' archeologist specialised in late bronze age to middle ages. She happens to be a horse rider too and a practioning fencer.
We are going to climb a local mountain this weekend and will be discussing merovinger swords and the like for hoooooooúrs :-)

Peter
http://users.telenet.be/huertecilla

One can talk about "historical" in rather loose terms or meaning something quite specific.
"Historically inspired" is the correct way to describe what usually is called "historically accurate" on todays market.
By using the word "historical" we must mean to compare in some way to something that actually did exist sometime in the past.
How much we aspire to compare will always be limited of what we know.
By using the words "historically accurate" we must aspire to somehting quite exact, or we denude the words of their meaning completely.

Historical swords are what they are because of crucial functional spects. The most important aspects are not even remarked on in the popular typologies, used today to classify swords. You need to see these swords first hand to get an idea of this. Or even better: get to handle them (a sword might give you one idea of its character from its looks, just to become a completely different beast when you get to hold it).
Just because a contemporary sword can be sorted and classified according to Oakeshott or Geibig, does not mean that it in any meaningful way is "historically accurate". It might look like a "typical" historical sword, it can have the general size, mass and perhaps even point of balance of the historical swords it aspires to replicate (even though all these vary). But it can still lack the actual character of the ancient swords.
So how could we be sure we hit the mark with all the variations given? How can we know what combination and set of variations are within the type? How can we say anything about what is atypical, if published texts and typologies does not even touch upon the nuts and bolts of the design principles?
-Only, by doing first hand studies of actual historical swords! Knowing one sword will tell us the specifics of one specimen, knowing a hundred swords will tell us something of the principles for thher design.

"Historically accurate" must also be specified as to what aspects it is meant to cover. Do we talk about material and making? Do we talk about style and looks, or do we talk about perfomance, handling and functon? Do we aspire to all these things at once? How much can we honestly try to cover? What aspects do we actually know enough about for our work to aspire to being "accurate"?
To some extent these aspects overlap, but not completely. You can make a focused study of a few aspects of the historical sword and happily leave the rest without regard. As long as the focus of the study is clear we can still talk about historical accracy.
It is common that you hear that a blade is forged with accurate tools and materials, following some hopefully accurate manyfacturing technique. The resulting sword can (and will) often look quite different from actual historical swords. In those cases the maker is interested in the craft and the materials. That is a worthwhile effort, but there is no guarantee that the resulting sword is in anyway similar to a historical sword, if it does not share the physical dimensions and dynamic properties of an original. Such a swords might be accurate in materials and techniques, but nothing else.
How a sword functions is primarily not about materials or manufacturing techniques; it is about design and dynamic aspects.

If you aspire to know about hitorical swords and be able to have something to say about what is accurate, you benefit by a personal study of actual historical swords. If knowledge is based on what is covered in publications on historical swords and swordsmanship, you are relying on second hand information at best and never know what is the prejudice and specualtion of the writer or what is actual facts. This holds true for both makers and users. Whithout first hand knowledge of and experience of histroical swords, the terms used to describe them then become quite shallow.
If you base your knowledge primaily on what is availeble on the interent, you are unfortunately on very shaky ground.
If you take part in discussions or make swords where the quality of historical accuracy is in focus, it is good if you name the source of your infomation to the benefit of your fellow enthusiasts.

Making a "copy" of a historical sword based on data limited to length, weight and perhaps point of balance, does not really come close to something that could be labelled "historical accuracy": there is far too much variation possible with only this information to build on. There is *no* guarante the "replica" will come even close to the original in performance or handling. Even if they might look somewhat similar when comparing photos, they could be two completely different swords seen side by side, when you can appreciate the difference in heft and dynamic balance.
The most important aspects for handling and performance in a sword cannot be described simply by length, weight and point of balance.

When you make up your mind about the "accuracy" of a design, you must consider where the maker got his data. A second hand source? Text in books? The internet? Generally acepted ideas among contemporary word afficinadoes?
Personal research on actual originals?
The outcome will be greatly influenced by this.
By saying something is accurate, it must mean that it can relate to something in a specific way.
If you allow it to be a very general and relative term, it becomes as meaningfull as "battle ready".

By saying all this I do *not* mean that there is only one way to make swords: by apsiring to historical accuracy.
I do think that both makers and users would greatly benefit from personal study of originals, regardless what the objective of the interest is. It would help a better understanding of the nature of the sword in its multitude of forms.

A badly made sword based on questionable research does not become better just because it is called "historically accurate".
A crooked length of dirty steel does not become a brilliant historically accurate scramaseax, just because it was made from bloomery iron and forged in a charcoal hearth.
It is possible to make an absolutely excellent contemporary sword without it having to be in anyway historically accurate. It is diminishing both for the maker and the understanding of the work if that sword should be labelled as such, however. Also, by making the distinctions blurry, we do not help our understanding grow, but instead get bogged down in the murky mud of shallow thinking and prejudice.

The sad thing is that "Historically accurate" has become a marketing term used to increase the allure of products that are anything but histroically accurate It confuses things, to put it mildly.

In those rare few cases "historically accurate" actually means what it says it will still only mean this: the object relates in certain specified ways to original historical objects. Wether this is something that is to be appreciated or not depends on your interests. If you like to learn something about the reality of the sword when it was actually used: accuracy towards historical examples would seem to be of great value.
If you are fascinated in a general way in long sharp blades, it can matter little.

News flash!

New evidence has just been discovered throwing renewed doubt upon what we all were led to believe were "Swords." You may recall the recent April fool's "joke" that there really was no historical artifact known as a "sword".

Or DO you? After forcing Albion to pretend that their expose' was really only an April fools' joke, the conspirators quietly removed any trace of the offending article!

However, recent photographic evidence proves that "swords" were really two-man dowsing devices, designed to identify and locate large deposits of water. (see below).

Unfortunately, these dowsing devices (there were thousands of them made) had a fatal flaw: they were weak where the two halves joined, and they tended to break as one man tried to go to the low spot in the field and the other kept turning around to watch the milk maids perform their chores. After one-man dowsing with a forked wooden stick became popular and the fragile (and expensive) metal dowsing devices fell into disuse and were subsequently forgotten, it was a simple matter for the conspirators to file the broken dowsing devices down and call them "Swords." The peaceful populace was all too eager to be seduced by tales of dashing "Knights" brandishing what were, in reality, harmless farm implements.

However, a rare production catalog was recently uncovered in the Vatican's water treatment plant which clearly shows the original shape of what has become known as the "Brescia."

This just goes to show that it is very difficult to know what these strange devices really were, or what they looked like in antiquity! Please don't be deceived, my friends! Even I can't tell a dowsing device from a fishing rod.


:D

edited to remove extraneous serious content

---

 Attachment: 17.16 KB

Two-man dowsing device: 299 Lire

I think it is probably worth noting that the extant surviving examples of dark age and medieval weapons represent a minute fraction of the total produced. Making definite statements about the average behavior and appearance of a weapon based on such an impoverished number of samples for many weapons (or style of weapons) is statistically invalid -- perhaps even foolish.

Take Viking age swords as a class example. While we know they were most definitely not manufactured from modern high performance steels, nor were they produced through the stock removal method, or CNC machined. I would challenge anyone to prove that anything is either true or untrue beyond that.

Let us for arguement sake say that over the course of the Viking age (which lasted several hundred years) , 25,000 swords were produced. Of that total number, let us imaging that 150 badly corroded originals survive to this day. That sample would represent 0.004% of the total -- that is four one-thousandths of the original number. Let us now imagine that the total number of original swords were produced with subtly different blade forms and pomel styles (as were the originals) with some featuring pattern welded blades and the rest without; the sample becomes even more divided and less representative of the class as a whole.

The upshot of all of this, at least in my view, is that we would be unwise to make any emphatic statement about how a certain piece should (or should not) look, handle , feel, cut, ballance etc. unless the sample size that generated that average upon which the statement is based is scientifically and statistically valid. Apart from pieces that seek to directly copy a singular know example in minutia -- and that is assuming that the original is in good eneough condition to even determine its original appearance and characteristics, we should only talk in terms of a reproduction's fit within the parameters set by the extremes in variation of the extant known originals -- and nothing more.

A.B.

Hellow everyone,

Last April, Drake Abrams started a thread on the Off Topic board in which he asked “What is “historical accuracy””. A number of interesting responses were tendered, but aside from my recent response, the thread seems to have died off last May. In my view, that is unfortunate because Mr. Abram’s question is a very important one to our community. It is for that reason that I have decided to re-pose the question here and to share the approach I would take in making such a determination.

My personal area of interest within the field of sword collecting centers on Viking Age reproductions, so I will use them as a class example to illustrate my position on this question – and I have no doubt that some of you will take umbrage with it.

First of all, we must clearly determine the criteria upon which any assessment is to be based. If a piece purports to be a “faithful copy” of a singular historical piece, any question of accuracy can be answered emphatically through direct measurement and analysis – simple as that. The piece will either be perfect (indistinguishable from the original at all scales of examination), or it will be less than so. The degree of departure from the original can be clearly determined mathematically. It must also be remembered that such an analysis can only be made on the basis of what remains of the original sword after centuries in the ground (or under water). If the piece is in anything less than pristine in condition we are forced to make assumptions about both its original performance and appearance, and those assumptions should be clearly disclosed. Assumptions always introduce a degree of uncertainty – the more assumptions one is forced to make, the greater the potential magnitude of induced error and the less definitive we can be.

If on the other hand the question being asked is “is this viking age reproduction historically faithful to the style and period it purports to represent", counterintuitively, the question may be even more difficult to answer. Principaly, the problem centers on the distinct possibility that the number of surviving original examples is too small with respect total produced to generate a reliable averages. If the sample size is further divided to accommodate Oakshot or Peterson typologies, the problem is amplified and further exacerbated by the fact that we have no idea of: a) how many swords were actually produced during the Viking Age, and by extension, b) how many swords within each of the variation categories (typologies) were produced. In short, we have no totals to work against, and therefore, no clear idea of how good the existing sample size is with respect to that total produced. Again, we must make educated guesses based on assumptions.

Evaluating the accuracy of any reproduction weapon should ideally turn on the degree of variation within each measurement category with respect to the measurements of a theoretical weapon (model) that mathematically represents the mean (average) of its type. Those familiar with multivariate statistics will recognize that these determinations can be made through ANOVA (analysis of variance). For the non-mathematically minded among us, ANOVA is simply a statistical process for separating the variability of a group of measurements into assignable causes and determining the significance of those variations with respect to the mean (average). I am unaware of such an analysis having ever been conducted. If such an analysis has been conducted (particularly for Viking Age swords) I would be most eager to see the paper in which the findings were reported.

ANOVA (and certain other statistical tests) would allow us to create a mathematical model of a theoretical sword that represents the statistical average of its type. It then allows us to compare any blade – be it of recent origin or antique – to that theoretical model and determine two things: 1) how much the sword being compared to the model varies from it (across and within all measured variables), and 2) whether or not those variations are statistically significant. In effect, such an analysis would completely remove subjectivity from the process.

However for such an analysis to be undertaken, several issues must be resolved – and a few might even be impossible to deal with. Again, using Viking Age blades as our example, it must first be determined if the number of extant original swords is large enough and varied enough (with respect to assumed total produced) to generate a reliable model of the average sword. If the sample size is too small, we are dead in the water as the probability of error in any result would be unacceptably high. Another concern centers on the condition of the Viking Age swords that do exist. Most are corroded, many show varying degrees of physical damage, and some only exist in fragmentary form. Because of metal loss through corrosion or damage some measurements (i.e. exact dimensional measurements and weight) my have to be extrapolated. Performance criteria would also be difficult to generate – most museums or private collectors would, for obvious reasons, balk at the suggestion of testing the cutting power or edge retention properties of their antique weapons directly. Needless to say, destructive testing is out of the question as well. We would likely have to rely on metallurgical analysis and lab testing to fill in the blanks. That having been said, if such an approach were feasable, it would certainly go a long way towards removing subjective opinion from the equasion.

In my view, determinations of historical accuracy should be made on the basis of standard scientific modeling techniques and statistical analysis capable of yielding a reproduction’s score with respect to the theoretical average original weapon of its type. Because of the unknowns and assumptions, I would set the threshold for significance (alpha) at 0.05 or 5% (which is actually a quite low). What that means is that if a reproduction varied by more than 5% from the theoretical average of its type, it would be significantly different, and therefore an inaccurate reproduction.

Just my thoughts on the issue, and I welcome your comments.

AB

Andy,
You're correct in that we don't know what the relationship of surviving swords is to the number originally made. However, you seem to be claiming that we can't say a sword is accurate unless we know it's not an abnormal example. A historic example is a historic example, regardless of its deviation from the idea of an unknown norm. An accurate reproduction of said sword is still "historically accurate" as it's an accurate recreation of a historic sword.

What we can't always say with certainty, which is perhaps what you meant to imply, is that a sword was "typical" or "normal" for its era.

A sword can be faithful to the style and period it represents if it replicates an extant sample. Period. That sample was accurate and representative of something in its period, even if it was out of the norm. Let's say someone discovers a sword with a scent-stopper pommel that can be irrefutably dated (this is the key to my hypothetical statement) to the Viking Era. Any faithful copy of it would be historically accurate and would represent both the era and sword it replicated. However, no one should claim that it represented a typical example, just the example it recreates. However, since it would indeed be a faithful copy of an historic sword, can we say it's not historically accurate? Id' say no. We can say it's likely historically atypical and leave it at that.

"Historically accurate" and "historically typical (or average or common)" aren't the same thing. For example, we know that there are a handful of great swords that were discovered in Viking graves. Any accurate recreation would be historically accurate, though no one should claim that Viking great swords were typical or common. Another example: brass/latten/copper have been used on sword and dagger fittings. So a recreation using those materials might not be inaccurate but likely wouldn't be typical or average.

If one is making a living history or educational kit/persona, they might want to stick with what is currently considered typical.

Most historical works are written and read with the implied knowledge that we don't know everything about history. After all, we might be one discovery away from having to re-write it all. So we can't say with absolute certainty what was typical but we can say what we believe is accurate and/or typical based on current knowledge and understanding. If every historian began or ended every sentence with "according to what we currently know," books would be hard to read indeed. :) We have to make logical assumptions based on current knowledge, while admitting the possibility that we don't know it all. As you said, we have to factor in sound assumptions to any equation.

Saying with certainty that something can't be typical is a unsound as saying it must be. So while we can't say we know what was typical with certainty, we can say what we believe was typical based on current research. That will have to suffice until we learn something that confirms or refutes what we know.

But, as I said before just because something may not be typical doesn't mean it's historically inaccurate.

Hi Chad,

Chad Arnow wrote:

Andy, You're correct in that we don't know what the relationship of surviving swords is to the number originally made. However, you seem to be claiming that we can't say a sword is accurate unless we know it's not an abnormal example. A historic example is a historic example, regardless of its deviation from the idea of an unknown norm. An accurate reproduction of said sword is still "historically accurate" as it's an accurate recreation of a historic sword.

I am not implying that at all. What I am arguing is that a reproduction that is, from a statistical perspective, significantly similar to the average of its type can be judged to be accurate. I also stated that a reproduction that is based on a known original -- no matter how unique -- can be judged by direct comparison of measurements.

It is, for example, at least plausible, that a blade of Persian origin might one day be unearthed in a Viking grave in Sweden. However, such a blade would certainly not be considered a Norse weapon and I would have to take a hard swallow it a maker created a reproduction of that blade -- no matter how accurate -- and advertised it as a Viking sword. It would deviate from the statistical norm by such a magnitude that no other reasonable inference could be made. The overwhelming probability is that it represents a unique item imported by a single individual. Now, I could of course be wrong -- there could have been tens of thousands of such blades in use and even manufactured in Scandinavia of this type but with only one surviving example ever having come to light -- but the probability of that being the case is statistically so low as to render the point moot.

My goal was to offer a methodology that would remove or signigicantly diminish the problems associated with subjective assesment and nothing beyond that.

Andy B.

Archaeologists use a statistics-based series of measurements and ratios of measurements to sort, type and date projectile points; a similar system for swords might not be necessary (for dating them), but if you could get a big enough grant it would be a really fun (for statistics) study to do.
;)