Class 1b/2 mules and class 2a, 2b: look for NN rather than the common HN.
HN coins become scarcer from about 3ab1 onwards, and examples should be sought, particularly for classes 3ab2, 3d2, 5a2, 5a3, 5b2, 5c1, 5c2, 5e, 5f (non-reported) and 5g.
Amongst 2b1 through 3b in particular, also 3d, look for HH coins, and particularly look for these in early class 5 (5b2).
3a1 through 3bc is where we begin to see pelleted “N”s beginning to appear – they reach over 15% for class 3ab2, but for other pre-class 3c coins pelleted “N” represents less than 5-10% of the total, and examples should be sought.
The combination of an “H” with a pelleted N is generally scarce – look for specimens in classes 3c onwards. Even scarcer is the combination of a pelleted “H” with a pelleted “N” – this has only been reported for class 5b2.
All 5d coins are scarce, but the ones with non-pelleted “H” or “N” more so.
The graph shows the fraction of the different mint spelling types, specifically the form of the letter “N” in the third and fourth reverse quadrants. Coins of the CAN/TER type are excluded, with the coins being portrayed being the ones with an “N” in both quadrants. It is readily apparent that there has been an evolution through time with HN initially being dominant, followed by NN (with no pellets) being the dominant type during class 3. “N”s with pellets are shown in shades of purple – these coins start to become important from class 3c and are the dominant type from 3d1 onwards, with the exception of the unusual class 5d and the immediately following scarce class 5e.
If the different combinations of “N” types were to have been found in a number of approximately equally sized groups, and that number of groups stayed fairly consistent from one class to another, then we might suspect variations related to die-cutter signatures, This is clearly not the case, and one may conclude that stylistic variation, fashion and/or die-cutter whim are the factors responsible for the observed pattern.
This small study has a side benefit of highlighting some scarcer “N” combination types that may be of interest to collectors:
Summary: The question of coded signatures comes up occasionally, and this short article addresses the topic and concludes that systematic “signing” by die cutters is unlikely to have occurred, based on a study of the mint name as portrayed on the coins of Canterbury. Unofficial die-cutter signature marks, particularly on the obverse of coins, where more variation is seen, cannot be ruled out.
First we might consider why die-cutters should need to “sign” their work – all the dies were cut in London and then distributed to the various mints around the country. Of course there was a need to keep a check on the minting process itself, and that is why the moneyers’ names are given on the reverse – they were responsible for maintaining the correct weight standard. But why should the authorities have needed to be able to keep track of the individual die-cutters work after the dies left London for the mints? So long as all dies were inspected before being approved for issue there would be no need for a die-cutters signature. Inspection was likely to have occurred and so there was no obvious business need for a systematic signing process.
Dies for the obverse and reverse may have been prepared by different die cutters, especially as the obverse and reverse had a different life span, and so more of one than the other would need preparing. Different die-cutters for obverse and reverse is supported by Ian Heaviside’s’ examination of class 3a1 coins. Hence if die cutters were required to sign their work then both the obverse and reverse dies would logically need “signing”.
Thus if one can demonstrate that the reverses were not systematically signed one might deduce that the obverses were likely not systematically signed either. Obverses do contain more variation in terms of various markings, pellets and stops etc. that have not been fully explained and could represent unofficial die-cutters signatures, and this possibility cannot be ruled out.
Turning our attention to the reverses, the design is very simple and variation is primarily through lettering variations including the use of H for N, ligations, pellets on letters etc. As Churchill and Thomas have usefully done a comprehensive job of fully documenting the reverse lettering of the Brussels Hoard coins they examined, one might be able to use their published data to address the signing question. They do differentiate between some differences in individual letter style in characterising various sub-classes, but they do not give as much detail of letter variations within any one sub-class, with the occasional exception e.g. usefully noting short cross letter styles. In other words there may be subtle variations in letter styling not recorded by C&T that we cannot analyse in a study of their published data.
We don’t know exactly how many die cutters were working in London at any one time, but one might assume that for much of class 3 and class 5 production there were no abrupt changes in the number of die cutters working. There likely would have been significant changes at the ramp up in production at the beginning of the provincial phase and a consequent downturn in numbers at the end of the provincial phase. Additionally, there may have been significant disruption at the time of minting 5d/5e coins.
With several periods of relative stability of the die cutters, by looking at the C&T data for provincial phase coins of classes 3a to 3c, and also Class 5 coins of 5a-5c, one can look for patterns of lettering variations with some degree of consistency from class to class. For example, if we can deduce in a class having a large number of samples say six different main groups of lettering style, of approximately equal frequency, then we might look for the same six groups in the class just before and after. This might be evidence for six different die cutters at work. But how can we measure the frequency of occurrence? We have a major problem in that C&T’s different types do not record individual dies but simply differences in reverse legends. Some of these types will consist of coins of just one die but others will consist of coins from different dies having the same spelling. We have no way of differentiating multi-die types from single die types.
A further complication in the analysis comes with the fact that one cannot expect equally sized groups from the same mint, as at any one time one die-cutter might be working on dies for one particular mint, and his colleagues working on dies for various other mints. It is thus easier to commence any analysis in the post-provincial phase where the coin production was from a more limited number of mints. I thus considered initially looking at the Nicole coins of London and Canterbury in class 5b2 and then compare to 5a2 and 5c1 coins. If there was any die-cutter signing activity going on then we should be able to detect signs of it in this sample.
Given that there were multiple moneyers per mint it may be more likely that any signature is concealed in the mint quadrants (3/4) rather than the moneyer quadrants (1/2), and so it is the last two quadrants that may conceal any signature.
Any such signature would probably need to be different for each mint, and so in London a record would need to have been kept of each die cutters signature markings for each mint; no such record has been reported.
The issue was further investigated through consideration of the Brussels Hoard coins from Canterbury. The majority of the coins there have a reverse text in the form of [Moneyer] ONC/ANT or similar. There are in addition a considerable quantity of coins in the form of ION/ON/CAN/TER or similar, and these have been looked at separately, with the main focus being on the coins of the moneyers other than ION.
To comment on or discuss this research note please visit the henry3.com blog.
ROBERT PAGE, Feb 19th 2015
- Systematic die-cutter signing is unlikely to have occurred due to the lack of a business need.
- No evidence has been found for die-cutter signatures related to H/N on the reverse of Canterbury coins.
- Unofficial die-cutter signature marks, particularly on the obverse of coins, where more variation is seen, cannot be ruled out.
- Some H/N combination types are scarcer than others.
Some examples are shown in the diagram above, which also serves to explain the nomenclature scheme on the graph below showing the changing fraction of the different mint spellings through most of the voided long cross coinage.
The above described scenario would strongly suggest that coded signatures were not being denoted by the use of pellets during the provincial phase, nor, of course, in the pre-provincial phase either. In fact, whatever the scenario regarding a changing population of die cutters, the low pellet usage through to class 3bc is evidence against coded die signatures by means of pellets during this period.
This leaves us with the question as to whether pellets were being used as signatures in the post provincial phase. Before addressing this in detail though, I will make the observation that the use of pellets peaked at the same time as we see the usage of other style changes related to finer quality die cutting. For example, the usage of “N” rather than “H” as described in research note #3, and the count of the number of pellets in the reverse inner circle (RIC) – which will be addressed in a future article. The data is thus suggesting that the usage of a pellet on the N is simply a matter of style, and that when die-cutters were taking more care in their work then pellets often appeared.
However, why is the usage not consistent? In other words, with two “N”s in the last two quadrants of the reverse legend, why is it that often only one of the “N” s has a pellet and the other not?
Looking at this above chart we can ask whether it tells us anything about the use of coded die-cutter signatures versus simple changes in style. The clear difference in pellet usage between the provincial and post-provincial phases might be explained by there being extra die-cutters for the mass production of the provincial period, with some of these die cutters no longer employed during the post provincial phase. In such a case if the longer term die-cutters were in the habit of using pellets and the temporary die cutters not, then this might explain the observed difference between the two phases.
The chart below clearly shows that pelleted “N”s were not a feature of class 2 and were scarce in early class 3, increasing rapidly from class 3c onwards – i.e. from about near the end of the provincial period. It may be interesting to compare the use of pellets from some of the provincial mints to the major mints of London and Canterbury. The use of pellets on “N”s reaches it’s peak of around 75% in in the 3d1-5c1 range, with the fraction in the latter part of class 5 falling back to around 50%, with two exceptions: no pellet usage in the enigmatic class 5d, and extensive pellet usage of almost 80% in class 5f. Keep in mind that these are the results from Canterbury and other mints may show significant differences.
The method involved examination of the reverse legends as published by Churchill and Thomas (2012), looking at the letter N in quadrants 3 and 4. Legends for over fifteen thousand coins of two dozen classes were analysed.
The results for classes with lesser samples, e.g. class 2a and class 4ab, are not as reliable as those from classes with a larger number of samples.
Before looking at the detailed results, we’ll first consider a chart simply showing the changing fraction of coins with pelleted "N"s through time. This next chart includes both the ON/CAN and CAN/TER mint spellings.