Bret Devereaux

...who assumed their experience of a given kind of warfare in the 1800s equipped them to understand the same 'sort' of warfare in the 1500s or 1000s. They were quite frequently very badly wrong. Assumptions about the wielding and weight of weapons...

...tactics in fencing and even the handling of cavalry all got built on that assumption that they were doing it how it had always been done. Again that is not to say experience with horses is useless, but that it is not the only kind of evidence or even necessarily the best kind

Consider, after all, sport rapier and sabre fencing can make all of the same claims of lineage that modern equestrianism can. They too are modern sport variants inspired by military arts which passed out of common military use in the first decades of the 20th century.

But compare them to the actual late medieval treatises and you realize that modern sport and ancient martial art do not necessarily coincide.

And of course sabre fencing as taught in 1910 as a combat art for officers and cavalrymen was not the same as fencing as learned by a 15th century man-at-arms.

Again, don't mistake me: experimental archaeology and modern practice have an important role to play in illuminating ancient or medieval practice. But it is not the paramount role: that role goes to the primary sources themselves.

Anyway, as far as experience goes, the last cavalry-on-cavalry engagement (the subject, I think, of discussion), the last was in 1920, during the Polish-Soviet War. There are, to my knowledge, no living veterans of that battle (given that if they were 18 then, they'd be 119 now)

Cavalry on *infantry* engagements have happened far more recently (there's been at least one in the current Afghanistan war, for instance), though they been mostly battles of dragoons, not lancers and their tactics do not resemble those of shock cavalry.