This year I decided to make multiple successive passes through a single target - and had multiple targets to eliminate. I first selected the one decorated in the lovely "barren oak tree" motif by my wife, shown here next to the same sword as last year.
I made several test passes in front of the pumpkin before moving in with offensive strike no. 4, per the inset below (fig. 3). Traditional sword strikes with the katana are designed to target soft tissues with large blood vessels, bleeding out the opponent fairly quickly. The are not designed for militant action against insurgent squash.
I failed to bring my pass down into the body of the gourd, instead taking off a thin portion of skin on the upper right and the majority of the stem. The next two passes encountered more success, but revealed a disturbing trend: the lower I cut, the more I favored the no. 6 strike.
This could be considered natural - after all, for right-handed katana use the no. 1, 2, and 6 strikes are the strongest. One may also observe within this diagram that between the no. 4 (neck: carotid arteries) and no. 5 (abdominal wall, intestines, descending aorta) strikes there is no strictly lateral movement: the bones of the arms and ribs make the gesture futile. Thus, time is not spent attacking in that attitude at intervening elevations and the swordsman is likely to drift into more familiar territory instead.
Said pumpkin was also sitting on a wall composed of cement. The further one descends into the no. 6 position, the more likely they are to encounter this wall. I'm not sure if I was too focused on the target to notice, or incorrectly believed myself to have accounted for the obstacle and adjusted my swing accordingly. Whatever the case, the fourth pass met with an unsettling CLANG- twing- *THUNK* -clatter-clatter and I became well acquainted with the interior handle construction of my sword - the blade of which was lodged 25' (7.6m) to my left protruding from some plastic buckets near the sandbox (the edged still remarkably intact, only lightly dulled and still formidable).
The tang reduces until it's screwed into a weight which was then epoxied into the pommel. That mount provided the tension to keep the rest of the stack together, and that's it. Everything else up to the tsuba/hand guard is cylindrically hollow allowing free movement - which may have done well to absorb shock, but drastically reduced strength and stability.
Well illustrated in the lower inset is the abysmal tang itself, measuring less than a ¼ the total blade width and breaking rather predictably exactly at that point.
I'm taking what's left of the blade and will mount that in the end of a bo staff for a home-built naginata. The handle remnants will probably be reassembled for use as a costume piece (snazzy lightsaber, or affixed to the scabbard as a non-functional safe-to-carry-anywhere traditional blade). The real difficulty will be convincing my wife to let me cut anything in the yard ever again.