Postbag in this months KM (Issue 96) Percy Blandford describes the Middlemans knot and its serious fault as a loopknot of being able to form into a slipping variant with noose like abilities..
Percy showed a method of tying which I thought I recognised as a method for tying the Alpine Butterfly, and when I tied the Middleman, sure enough I had the Alpine, but after just a little adjustment it slipped apart into two slipping OH knots exactly as Percy had described.
Had the Alpine always suffered from this potential to ensnare its user?
I could see immediately that the likeness was broken by one tiny fact – the two overhand loops were enmeshed in the Alpine Butterfly but were not enmeshed in the Middlemans knot.
I then tied the Alpine using Dave Roots method No. 1 which is the method I had confused with the one described by Percy. In the Alpine method, the loop is given two twists and the lines cross over then under one another - naturally it finished up as the Alpine (enmeshed OH loops).
When I looked again at the Middlemans method, although the shape of the ‘C’ illustration is the same as the Alpine intermediate shape, the difference is that in the Middleman, one cord is simply laid over the other – it crosses, then uncrosses, whereas the Alpine crosses then crosses again. That tiny difference, a cross under instead of a cross over makes an Alpine or a Middleman.
I think I will stick to the hand loop method in future.
I responded to Percy’s article in Knot News (PAB’s newsletter) in which it earlier appeared,
pointing out the shortsightedness of knotters re that, possibly induced by tying method:
adding one more Overhand can eliminate the noose aspect (and this Overhand can be
put on at either end of the finished offset Fisherman’s knot loopknot). This structure can
be seen in Jost Gudelius’s presentation of the corresponding offset bend:
By coincidence, I’ve just discovered midline LK versions correlating to Ashley’s #1408
(which I’ve long regarded as a pure (i.e., symmetric) version of the Butterfly) and
then by self-direction (i.e., telling myself, “If that exists, there should be one …”) for #1452.
This latter should better resist jamming, and in its asymmetric nature can be regarded as
a hybrid of Carrick & #1452 structure–just as the latter can be seen as a strengthening/securing
of the former–; the direct recipe used for #1408 didn’t result in a Tiable-In-the-Bight (TIB) knot,
and my workaround changed one Overhand into a Crossing knot, hence the Carrick half. ***edit/update: Seems my enthusiasm overwhelmed my comprehension: the #1452 knot
indicated in fact isn’t TIB–argh. An alternative with #1452-ish characteristics HAS been
found, now, with not a Crossing-knot but a Fig.8 in its place; it looks good.***
There are symmetric, twin-eye versions of 1408 & 1452 (and …), where the point is
to load the eyes jointly qua one, and thus that part that would’ve collapsed the collar of
the unloaded SPart (if indeed unloaded–depends on use of the knot) is not free to do
so because the twin eye is rooted around the loaded SPart and not free to extend so much.
(At least this is what it seems at some moderate loading I’ve tried.)
To answer Roo: (1) Derek “thought [he] recognised as a method for tying the … Butterfly”;
(2) no, he realized his mistake. Root’s 1st method is apparently a modification to one of the
more commonly presented tying methods (twist & twist again, then fold over … and bring bight
tip around & through twist holes) aiming to better hold open the insertion channel for the
bight tip; Derek’s flub of this raises the question of the method’s success in that (or of Derek’s
following instructions). Root might enhance his method-1 by casting the 2nd twist also around
the arm; but while this method makes bight insertion easy, it is tricky to make the correct follow-on
twist (as I just confirmed/echoed Derek’s flub–twice!).
