William Shepherd studied classics at Clare College, Cambridge, in the 1960s and then embarked on a career in publishing, which finally brought him to Osprey, retiring from the position of chief executive in 2007. He is author of The Persian War (Cambridge 1982) translated from Herodotus. He has also written reading books for children and articles in the Osprey Military Journal, of which he was joint editor, and makes regular contributions to this blog. He lives in the Cherwell Valley, north of Oxford.

Since leaving Osprey he has become an Osprey author. writing two campaign titles, Campaign 222 Salamis 480 BC and Campaign 239 Plataea 479 BC. His next Osprey book, Pylos and Sphacteria 425 BC is out in December this year and is available to pre-order now!

In this blog post, William talks about some of his research on ancient weaponry.


Ancient Weapons Research

The start of a new Persian War project and some comments in Amazon reviews on points of detail in my Plataea book have caused me to take a critical look at one or two “facts” that I confidently include in my descriptions of hoplite weaponry. 

For example, I say that vase paintings clearly show that the linothorax (linen cuirass) progressively superseded bronze body-armour from the latter part of the 6th century.  But it’s actually impossible to tell what material the painters were depicting and a strong case can be made that leather was much more widely used than linen (glued in layers) in place of bronze. Leather was the least costly, the most easily worked of the three (bearing in mind the process of manufacturing linen fabric) and, if less resilient, could easily be reinforced with metal scales, as “composite” cuirasses often were. To strengthen this case, the few mentions of linen armour I have so far been able to find in the literature (two in Homer, one in Herodotus and one in Xenophon, and there the noun, linothorax, does not appear anywhere) make it sound quite special and exceptional.

Then, on the doru (hoplite spear), I wrote “Practical experiments have demonstrated that a thrust delivered overarm was far more powerful than an underarm thrust”.  However, I’m had to admit I have not seen the evidence from these “practical experiments”; I’d just read about them somewhere, and the conclusion seemed to make sense…

And then, on the web this week, I met Lloyd , and about time too!  He makes some very good points about the hoplite cuirass in one of his many highly entertaining and well-informed video rants (“rants” is his own term for them).  He also argues very strongly in Great Weapons of the Ancient World number 24: the Spear that the over-arm hold was much less effective than the under-arm, even to the extent that it may rarely have been used.  I did find myself wondering if the depictions of it in vase paintings can be so easily dismissed as “for dramatic effect rather than for authenticity”. Also, at a practical level, if, as seems to be the case, the doru’s  point-of-balance was about two-thirds of the way from its head, its reach when used over-arm would have been rather better than argued.  Finally some of Lloyd’s argument against the over-arm theory is based on weight and it is possible that his replica dark-age 8-footer was heavier than the average doru. (Lloyd’s insights are so valuable because he seems to have worn or handled accurate replicas, which he has quite often also made himself, of almost everything he writes or rants about.)

Two pieces of research would be useful here: a survey of a decent sample of vase paintings depicting combat to quantify the proportion of over-arm to under-arm holds; and a survey of surviving spearheads and butt-spikes to establish the thickness of the shafts and to arrive at an average weight including 8ft of ash for the original weapon. Both may have been done, and in any case I must now track down that research on the power of spear-thrusts.

Finally, the penetrative power of Persian arrows – One reviewer challenged the assumption underlying this detail:



In his excellent book on Marathon Stephen Krentz cites research which shows that a Persian bow could generate maximum kinetic energy of 50 joules, which puts it on a par with the English longbow.  At a range of 40 metres or less this would be sufficient for an arrow to pierce a hoplite shield, which was found to require energy on impact of 25-35 joules. By the way, Krentz’s book is the best of three titles published in 2010/11 for  the 2,500th anniversary of this battle; the other two are also well worth reading.  Plataea’s anniversary comes up in 2022 so we got in early at Osprey!