Tuesday, 12 July 2016

Lies and Damned Lies: why the average size of an erect penis is getting smaller...

This has been a week for statistics. Actually, it has been a month for statistics. Just about everywhere you turn someone in a suit is throwing some statistic at you. Statistics, we are meant to believe are, if not solid facts, then the building blocks for truths. They are the foundation of our learning, the cornerstones of our studies. For me too this has been a week of statistics - medieval graffiti statistics - and I have come to hate and mistrust them. I have come to see them as the shallow things that they really are, designed to mislead, designed to delude. They are, if I'm honest, a bit of a bugger...

You see, the thing about statistics is that, used selectively, they can be used to back up and legitimise just about any argument you want. And I really do mean just about anything. Don't believe me? Well try this one for starters.

'It is the opinion of this house that the average size of an erect penis for a western European male is actually diminishing over time'.

First of all we have to start with the irrefutable facts. Today, according to a study published in 2016, the average size of an erect penis for a British male is 5.16 inches. That's right everyone. Nor six inches - but 5.16 inches. This figure appears to be a fairly constant, with another study published in the 1940s also coming up with a figure of between 5 and 5.3 inches. However, due to the benefits of a better health care system, better education amongst parents and far better levels of infant nutrition, the average overall physical size of the British male is increasing. In 1954 the average British male was 5ft 7inches tall, and weighed 11st 6lb. In 2016 that has increased to 5ft 9inches, with an average weight of 12st 6lb. So, in the intervening six decades the average British male has become 2 inches taller and a stone heavier in weight, but the size of their erect penis has remained roughly the same. Therefore, in terms of penis to body mass ratio, the average size of average erect British penis is proportionally smaller.

Now if I really wanted to add the air of authenticity to this argument I could fall back on the art historical evidence - very selectively obviously. First of all there is all the Roman art and artefacts - many showing certain Roman individuals with incredibly large erections - implausibly so in some cases - but who are we to argue with historical 'fact'. Then you can bring in a whole series of C18th engravings and prints, in which the average sized penis appears to be well over 9 inches. Bringing it even further up to date, you can show images of the amazing graffiti at Bempton army base, itself now a matter of historical record, which indicates that 8-10 inches was the average during the Second World War and immediate aftermath. No wonder Hitler lost...

Okay, okay... you will all have noticed by now, I hope, that the above argument is based entirely upon either the selective use of evidence, or the manipulation of certain statistics; statistics that were never really meant to be comparable in the first place. In a debate relating to penis size it is really pretty easy to spot the flaws in the argument. However, such misleading and manipulative use of statistics can be found around you just about every day - perhaps more so given the recent political climate. In the last few weeks I have heard both politicians and respected political commentators using terms such as 'the majority of the British population voted for Brexit'. Last week it was a senior politician on the BBC using those exact words.

Frankly they should all know better - and so should we. What they are TRYING to say is that the majority of the people who could be bothered, or were allowed, to vote in the recent referendum voted to leave the EU. Seventeen million of them. However, as any politician and political commentator should know, the population of the UK is approximately 64.1 million. So, all politics aside, and with the very best will in the world, I still can't get 17 million, as a proportion of 64.1 million, to be any more than about 27% - and that's allowing for my poor level of education in a state school obviously. Lies, damned lies - and statistics.

The problem of course is that the study of history, be it political history or archaeology, is utterly devoted to the use of statistics - even the study of medieval graffiti. Rarely a day goes past without me being stopped in the street and asked 'what proportion of ritual protection marks are made up of compass drawn designs'? Okay... maybe not - but you get the idea. So imagine my interest then, when an article landed on my desk last week dealing with medieval graffiti - an article that, based upon certain statistics, had drawn some very, very interesting conclusions. Obviously I'm not going to mention the author - because we have already discussed the matter at length, and the article is currently being heavily revised. However, it is worth looking at the two main conclusions of the original article.

1. That the majority of medieval church graffiti is found on the pillars and exposed stonework of areas such as door surrounds, thereby indicating a deliberate choice and placement of those areas by the authors of the graffiti. Such a choice of these harder stone areas strongly suggests that the author associated permanence with an enhanced potency in regard to devotional inscriptions.

2. In light of the findings of Easton and Champion there are strong indications that the early modern period saw a shift away from the creation of graffiti inscriptions in churches and towards vernacular buildings. There is little evidence for ritual protection markings in pre-reformation domestic properties and, as Easton suggests, these appear to be largely confined to the seventeenth and eighteenth centuries.

Both of these arguments were backed up with lots and lots of lovely data and statistics - some of which were actually drawn from this survey - and both conclusions appeared superficially reasonably compelling. Difficult to argue with. Except that, in reality, and exactly like my argument above, they are both complete cock!

In some respects the first conclusion is slightly more difficult to question, largely because it is entirely true. In 80%+ of churches this is exactly where we do find the graffiti - on the exposed stonework of pillars, door surrounds, chancel arches, etc. However, this is to take the known statistics in isolation. To just deal with the figures and completely ignore the wider picture - the context from which those statistics were drawn. And the context is this. In almost all churches, even to this day, the internal face of the walls was plastered. Today very little of this medieval plaster survives, particularly not in the lower areas of the walls; the areas in which we normally find graffiti inscriptions. However, at sites where medieval plaster does survive, such as in areas at Swannington or Troston, then these too are covered in early graffiti. Therefore, the distribution pattern that we see in the graffiti is a false one. It is NOT a distribution pattern of medieval graffiti, but a distribution pattern of 'surviving' medieval graffiti - and the two are very different things.

The second conclusion is, it must be said, on the face of it, based upon good data. We do find most ritual protection marks in a domestic setting on buildings that post date the reformation. Easton's conclusions ring true. However, once again, we are looking at data and statistics that are, at best, incomplete - and as a result just as misleading. Put it this way, if you are mainly looking at post-reformation buildings then that is where you are going to be drawing your data from. Medieval domestic buildings are far rarer survivals, and those that do survive have, like all older houses, suffered restoration and renovations beyond count. In essence, we simply don't have the surviving medieval housing stock available to survey in the same way that we do have from the post-reformation period. However, in the few cases we have been able to look at such structures, then ritual protection marks ARE present - but the overall number of sites examined to date means that these sites barely dent the statistics. They may be all we have - but statistically they don't really count.

So what then is the lesson from all this? Is there indeed a lesson at all? I suppose it is simply to urge caution. To always question every statistic that is thrown at you, whether to do with medieval graffiti or politics, and look more closely at both the data itself, and the context in which it was created. Statistics alone are rarely meaningful, particularly in relation to historical research. They are, quite simply, lies, damned lies, and statistics - and nine out of ten cats agree with me on this...

1 comment:

  1. Many years ago I had to do a statistics module as part of my degree course. One of the points I remember was that a result in statistics was not an end in itself but the starting point for fresh questions as to what did the result mean.

    The Radio 4 programme more or less is interesting for the examination of statistics that are widely used in the public realm, some of which are quite spurious, but having been widely quoted are accepted as fact.