Author Topic: Classification.  (Read 8109 times)

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1555
  • Knot Botherer
    • ALbion Alliance
Classification.
« on: January 05, 2016, 06:06:33 PM »


Until now, we have classified knots as Bends, Loops or hitches, i.e. we have classified by principle form of usage.

But what if we put less emphasis on use of a knot, and instead focussed on what the knot was made of and how it worked?

Knots are built up from identifiable components.  These components are assembled, 'wired up', in various ways to achieve a variety of functionalities.

Obviously, how the components are connected, will influence how forces are processed within the knot - these are after all, force machines...

Some knots have a clearly defined structure that is so effective and useful, that they have been reworked / re-purposed to cover a range of uses, i.e. as the heart of bends, loops and hitches.  But these self functioning structures, on examination, are in turn formed by the assembly of fundamental components, which by themselves are incapable of functioning as a working knot, yet crop up again and again in the vast array of knots that have been developed and documented throughout history.

Which components are used, and how they are connected, directly influences the form, function and usability of the final knot.  It is perhaps no surprise that there is a great parallel between components in knots, and atoms in chemical compounds, where, which atoms are used and how they are arranged directly influences the form and function of the final compound.

So, with this perspective in mind, let me start by suggesting a small set of the most frequently used components:-

LC - the Load Component.  Cord upon which a load is applied either into the knot or by the knot.

This component may seem a little irrelevant, but for all working knots (as distinct from decorative knots) the application and processing of force is the sole purpose of the knot.

TC - the turNip Component.  The simple helical nipping loop.  This deceptively simple component is highly versatile and will be discussed extensively in its own thread.  Mostly used with other components, but can be combined with itself as in the case of the Myrtle Knots.

BC - the Bight Component.  A 'U' turn of cord, often used as a collar.  Mostly used with other components, but can be combined with itself as in the case of the Square Knots.

CC - the Carrick Component.  Although this component can be described as being formed from two BCs, its valuable characteristics earn it a place as a component in its own right.

There is just one more component of significance to working knots, and that is the End Component or EC.  With the exception of TIB knots, every knot will have at least one of these.  Their security is of great importance.

These five basic components between them can be found in virtually every knot we know of.  Obviously every working knot must have at least two LCs, and the large majority of known knots will be formed by arrangements of one or two of the functional components.

Although we do not yet have a system, such as Chemists have, to depict how these components are connected, we can at least start to map our known knots according to the components they are made of.

For example, the Reef knot has two load components LC2, two bight components BC2 and two ends EC2, so it has  LC2, BC2, EC2.

Or the Carrick bend has two carrick components so is LC2, CC2, EC2.

The Bowline has an LC, then a TC and a BC and an EC.  But it also has a loop.  We can either conscript the Bight Component into use as the loop or create a new component as the loop.  If we use the BC then the Bowline would be mapped as LC3, BC2, TC, EC.  You will notice I mapped three Load components, this is to identify the two loop legs which will be loaded.

Anyway, that is what I have been working on for a while now - so, if you ignore the (silly) letters, what does everybody think about carving up out knotscape by components rather than by use?

Derek

KnotMe

  • Sr. Member
  • *****
  • Posts: 632
    • The Chinese Knotting Homepage
Re: Classification.
« Reply #1 on: January 05, 2016, 10:08:47 PM »
Potentially very useful, especially if it can be used to encode knot descriptions for computer manipulation, but at first glance extremely arcane and not very human readable (but then you knew that already  :D ).

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4083
Re: Classification.
« Reply #2 on: January 05, 2016, 10:27:25 PM »
Until now, we have classified knots as Bends, Loops or hitches,
i.e. we have classified by principle form of [use].
Whoa, there : don't "we" so freely --*I* don't have
such a simplistic classification (as this historical trio),
for by use there are binders & stoppers to add to
end-2-end knots, eye knots, & hitches.  (And I split
out, qua "knot structures", such things as nooses
& trucker's hitches --things comprising component
knots but themselves a different entity, IMO.)

As a way to set this out, just make a circle with an 'X'
in it and however many ends (multiples of 2, necessarily;
illustrate the connections via ID such as "1-2", "A-B")
are of issue, and then indicate the loading profile
(oh, one may include an object through this circle
--such as the noose's S.Part running through the
clove hitch that is known as "2 Half-Hitches"
and usually (the entirety) considered a "hitch" but is
what I call a "structure" (a noose)).

Quote
But what if we put less emphasis on use of a knot,
and instead focussed on what the knot was made of and how it worked?
I have mused about knot "types"/"kinds"/ (... such terms in
any way will have various senses --pic one...!),
with "interlocked", "traced", "pull-together" (think "grapevine"),
and so on.  --that's pointing to a more general mechanism
than the actual geometry of the interlocked/traced/pulled-together
components; but it's a way of distinguishing knots.


--dl*
====

agent_smith

  • Sr. Member
  • *****
  • Posts: 1391
Re: Classification.
« Reply #3 on: January 06, 2016, 01:25:43 AM »
I find it troubling that IGKT members cannot agree on individual knot component terminology. Why is this so?

(Sorry Derek...time for a quick rant          rant rant rant)

For example, I still see a myriad of names/terms given to the 'nipping structure' in a Bowline:
[ ] turNip
[ ] nipping turn
[ ] nipping loop
[ ] nipping helix
[ ] nipping component
[ ] nipping structure
[ ] helix
[ ] helical component
[ ] helical compression component
[ ] radial compression zone
[ ] radial component
[ ] radial containment zone
[ ] compression zone
[ ] add your personal favourite... (I used the online thesaurus to find a few extras!)

Surely it cant be that big a deal to sort this out? If we can figure out quantum mechanics and SpaceX can soft land a first stage booster (and re-launch it again) - why cant we sort out our own house?

and...

loop, eye, bight, wraps, 180 degree U turn, 360 degree turn...

////////////

I know that when I walk into a hardware store and ask for a 'loop bolt' - the salesperson will think I am 'loopy'. But if I instead ask for an 'eye bolt' - I should receive something that resembles the attached image below. Or if I go to a rigging company and ask them to make me a 'loop splice' - again they will think I am loopy. Or a loop sling instead of eye sling...I would need to say 'eye splice' or 'eye sling'. So what is the issue with using the descriptor 'eye knot' versus 'loop knot'? (or loop bolt versus eye bolt?)...

NOTE: Ring bolt is different to an eye bolt. See photo below. But its definitely not known as loop bolt or a bight bolt!

Sorry Derek, I think this will bog down because there is no universally accepted knotting terminology. I guess I'm saying that we first need to get our house in order before stepping outside.

And this all sounds negative doesn't it?!

Sorry Derek..you have made an interesting post and I like the concepts behind what you are doing. I'm sure this can all be investigated and expanded upon in detail.

Quote
Although we do not yet have a system, such as Chemists have, to depict how these components are connected, we can at least start to map our known knots according to the components they are made of.

Yes - a periodic table of knots

This has been talked about before (in 'Rope Test Lab' on facebook) but, it quickly bogged down and vaporized into the thin air.

I like the concept and would be pleased to assist - we do need to do this (sooner rather than later).

I can say that I have considered organising Bowlines according to their 'nipping structure' into a sort of 'periodic table of Bowlines'. I think this can be done in theory...

Quote
BC - the Bight Component.  A 'U' turn of cord, often used as a collar.

I had thought that collar structure might/could supersede and render the term 'bight' redundant?
« Last Edit: January 06, 2016, 04:15:48 AM by agent_smith »

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4083
Re: Classification.
« Reply #4 on: January 06, 2016, 02:18:53 AM »
I find it troubling that IGKT members cannot agree on individual knot component terminology.
Why is this so?
Because it's problematic and this isn't a group
of linguists or philosophers or scientists/engineers!?

Quote
For example, I still see a myriad of names/terms given to the 'nipping structure' in a Bowline:
[ ] turNip
[ ] nipping turn
[ ] nipping loop
[ ] nipping helix
[ ] nipping component
[ ] helix
[ ] helical component
[ ] helical compression component
[ ] radial compression zone
[ ] radial component
[ ] radial containment zone
[ ] compression zone

[ ] add your personal favourite...

Surely it cant be that big a deal to sort this out? If we can figure out quantum mechanics and SpaceX can soft land a first stage booster (and re-launch it again) - why cant we sort out our own house?
You must really be making hay to have so many
strawmen erected in just one post!  --the reddened
entries are IMO purely your creation and nothing
that has been troubling others.  "turNip" is purely
my cutesie/word-fusion moniker tossed in for quick
chatter, nothing I'd expect to travel up in First Class
seating.  (Btw, I've not figured out quantum mechanics,
or even many car mechanics.  As for "sorting my own
house," I've a long list of new-year's resolutions and
this usually tops it --annually.   ??? )

Quote
loop, eye, bight, wraps, 180 degree U turn, 360 degree turn...
"Eye" is my urging so to stand clear of "loop"
which is in the common tongue; "eye" should
be UNconfusing but for being different --i.e.,
it shouldn't carry ambiguity, and with the simple
association --pointed out, if need be-- to the
common "eye splice" be acceptable,
even though we cannot erase "loop" usage.

Whereas "bight" has problematic aspects in its
knotting uses ("without ends" vs. "doubled"/folded
in half"/"U-shape(d)"), and then its seemingly
forgotten nautical (& inland marine, riparian?)
sense of some sort of curvature.  (I myself find
it hard to speak of a rug's manifold "bights" vs.
"loops" of material.)
And then there's all of the 'bite's to fite, er, fight!

Similarly, "loop" is overloaded,
and "turn" as you indicate comes up for question
in trying to convert it into degrees --for sure
clarity, maybe best (word-fusion : 'mayBEst' !)
to attach a degree numeral to it, and let the
"turn" part indicate simply the rounding,
and the numeral specify the extent --not a
handy thing for teatime parlance, but quickly
and surely comprehended when encountered.


--dl*
====

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4083
Re: Classification.
« Reply #5 on: January 06, 2016, 02:29:18 AM »
To clarify (or just go on about, further) my prior
remarks, I see a simple graphical notation to set
out classifications --that circle & 'X' to simply
denote "knot(here)".  (I sometimes use this where
specific-knot imagery isn't needed --"whatever".)

Each strand running through a knot has of course
two ends, and I give ID to such strands as noted
in the above post as "1-2" & "A-B" and so on.
So ID'd, one can specify loading profiles as, e.g.,
"1 vs 2+A" for an eye knot (2 would be here what
Agent_Rant calls "the ongoing eye leg" and A the
"returning" one, leaving B to be tail.  If this were
for the bowline (#1010), then "2 vs. 1+A/B"
would give Eskimo bwl.s, one/other.
And so on ... .


--dl*
====

agent_smith

  • Sr. Member
  • *****
  • Posts: 1391
Re: Classification.
« Reply #6 on: January 06, 2016, 04:23:31 AM »
Quote
You must really be making hay to have so many
strawmen erected in just one post!  --the reddened
entries are IMO purely your creation and nothing
that has been troubling others.

I wish I could find even more permutations and combinations...if only to illustrate that the terminology/syntax is still to this day not agreed upon. eg
Quote
TC - the turNip Component.  The simple helical nipping loop.
here we see a number of descriptors being used. Again, I like what Derek proposes - I just think we need to lock down a few terms first (ie get our house in order!?).

... 'BC' (bight component) - could be 'collar structure' for instance. I had thought that there was an implied acceptance of 'collar structure' in place of bight component...?

And the 'ongoing eye leg' and 'returning eye leg' (per Bowlines) - I thought those descriptors had merit? Was it Xarax who first proposed those descriptors for the eye legs? I am not sure...

« Last Edit: January 06, 2016, 04:25:12 AM by agent_smith »

Sweeney

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 975
Re: Classification.
« Reply #7 on: January 06, 2016, 11:36:16 AM »
I find it troubling that IGKT members cannot agree on individual knot component terminology. Why is this so?

I think you mean "... a very small group of IGKT members...." - the vast majority are almost certainly living in ignorance of this discussion. My own view for what it's worth is that any method of coding (unless purely for internal use in a computer system and unseen by the user) is doomed from the outset - people just can't relate to it. The photos showing a ring bolt vs an eye bolt sum this up precisely - we can all recognise or point to a picture and that is where Ashley scores - so many knots do not have a name (or multiple names) but by using the reference number associated with the knot we can then trace and look at a picture so why worry about some abstruse coding methodology? Why not simply create a set of pictures of knot parts with a text description and a reference number? I don't care what a "nipping loop" is called as long as I can say "it's picture #37" or whatever. The trouble is that this takes a lot of work, support from the IGKT as an organisation and some defined benefit to the world of knotting which seems to be managing pretty well without it.

Sweeney

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1555
  • Knot Botherer
    • ALbion Alliance
Re: Classification.
« Reply #8 on: January 06, 2016, 11:42:15 AM »
Potentially very useful, especially if it can be used to encode knot descriptions for computer manipulation, but at first glance extremely arcane and not very human readable (but then you knew that already  :D ).

Carol,

Arcane is exactly what I did not want.  I believe that we need a system of knot description / understanding that can be the starting point for anyone interested in knotting.  Hiding that knowledge behind a wall of knottese would be worse than the ridiculous wall of jargon a new knotter hits when they first find out they must now call certain knots Bends, or tangle with Bights...

The 'man on the street' has access to our world the moment 'she' picks up a piece of cord, yet if they have any interest, the moment they look further they are assailed by a thousand different structures, often referred to by tens of thousands of trade or usage specific names and an amazing assortment of making methods.  Worse of all, it has no order and has to be committed to rote memory.  It is no surprise that people turn to Sellotape, glue, Velcro, or at best they use one or two easy to remember knots.

But what if we could produce a knotters equivalent of a set of Lego bricks?

As a child, my imagination was ignited when I was given a Mechano set, and today my four year old Granddaughter creates wonderful places with just a few types of Lego bricks.

What could young (and not so young) minds do if we showed them that those thousands of different knots were actually built up from a small handful of Knotto bricks (components).  Explain how these bricks lock together and the special properties each brick has, then let them build away.

Instead of thousands of discrete knots, they would only have to learn a dozen or so components and then the world of knots is theirs.

I understand Mark's frustration over terminology.  As a Chemist, I appreciate the importance of a common language, yet our human propensity to develop new words to name new compounds is amazing, and we still have gardeners using Sulphate of Iron and Printers using Acetone...

Yes, language is a problem, and I believe will remain so for a very long time.

I was deliberately verbose in my introduction of this concept, in the hope that using several descriptors, the reader would understand my meaning.  I even descended to tokenised symbology, but perhaps that was a step too far at this stage.

Perhaps I should have resorted to coloured images of the components and referred to them by their colour.

My objective here is to promote the concept of the Knotting equivalent of Lego bricks.  Honestly, I care not what we call these bricks, only that we extract from the brain aching mire of knotting history a simple system that allows us to play with knots and understand/ think about, how they work.

The concept is perfect for our modern computing power.  An app that has a pallet of components that can be dragged onto an assembly space and linked up, then add load components and have the program show the knot respond to load for you.  Although very doable, yet beyond my skills I am afraid.

Anyway - thoughts on how to progress this concept and issues that need resolving - here would be a good place to share them..

Derek
« Last Edit: January 06, 2016, 12:20:20 PM by DerekSmith »

Wed

  • Sr. Member
  • *****
  • Posts: 314
Re: Classification.
« Reply #9 on: January 06, 2016, 01:02:05 PM »
I see a likeness to this old thread: http://igkt.net/sm/index.php?topic=2209.0. It was a really interesting exercise. And it had some merit.

And it isn't the only time the same subject has come up. I'd like to be rid of the ambiguity that nomenclature of today presents. That more or less require deprecation of old terms. That in itself will probably be more painful than the US metrication process.

Add to that there are no (to my knowledge) academic institutions in any country to encourage or enforce any change. The IGKT is not either an academic body.

And then only English terms have been discussed. In Swedish for example, any reference to English knot terms just don't exist. Hitch or bend? They are just known as knots. So making a global set of definitions, may well prove challenging.

I have a fuzzy memory of reading about forensic knot theory regarding number of crossings. Maybe that is worth reading up on.

Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4083
Re: Classification.
« Reply #10 on: January 06, 2016, 05:17:42 PM »
The IGKT is not either an academic body.
//
So making a global set of definitions, may well prove challenging.

In ISO (the int. standards body --per English initials, not the French :-),
I think that knotting will fall under a Textiles (working?) group?
 (I at sometime previously made some weak effort to learn this,
  being then more engaged with things touching ISO standardization.)

IMO, there could be some sharply articulated technical language
that would suite knot-theorists, but be not so portable into the
common language(s) --but perhaps even there, the exercise
of articulating concepts might have some benefit.

 - - - - - - -

Now, back to ideas:
consider that the overhand "component" might be
used in --citing well-known knots-- ::
a traced manner as the water knot/ring bend & "EDK",
or in a pull-together way as in the fisherman's knot,
or interlocked as in Ashley's Bends #1452/1408/1425/1425a;
and it can also serve qua stopper knot in resistance.


--dl*
====
« Last Edit: January 08, 2016, 09:53:10 PM by Dan_Lehman »

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1555
  • Knot Botherer
    • ALbion Alliance
Re: Classification.
« Reply #11 on: January 08, 2016, 10:59:50 AM »
I see a likeness to this old thread: http://igkt.net/sm/index.php?topic=2209.0. It was a really interesting exercise. And it had some merit.

And it isn't the only time the same subject has come up. I'd like to be rid of the ambiguity that nomenclature of today presents. That more or less require deprecation of old terms. That in itself will probably be more painful than the US metrication process.

Add to that there are no (to my knowledge) academic institutions in any country to encourage or enforce any change. The IGKT is not either an academic body.

And then only English terms have been discussed. In Swedish for example, any reference to English knot terms just don't exist. Hitch or bend? They are just known as knots. So making a global set of definitions, may well prove challenging.

I have a fuzzy memory of reading about forensic knot theory regarding number of crossings. Maybe that is worth reading up on.

Hi Wed,

Yes, I have been pushing various ideas around for years now, starting with the Overs Index way way back.  But everything had its own serious limitations, and behind it all was the quicksands of nomenclature.  It swallowed every attempt at standardisation with nothing more than a burp and a chuckle of satisfaction.  Rather like the sewage pond in the Blaster Bates story he titled 'A shower of Sheiet over Shropshire'.

Well, just like Batesy, I am going to try shoving a few sticks of jelly into out favourite sport.  I can't beat the nomenclature problem, so I am going to try stepping around it (a classical Dunnyman ability).

If we can at least begin to start the process of classifying our knots by their key components, then the naming can follow (if it wants to...).

So, in line with my previous post, I am producing a set of descriptions and images of coloured knotting Lego bricks.  Then we can all have a play and see how far the idea takes us.  There will be problems, knot structures that just don't fit the simplistic Lego brick concept.  We will have to either put them in the 'Other' Group (AKA 'too hard to get my brain around') or try to identify and describe new components we come across.

The biggest problem I foresee, is that this is only a partial solution, because although we will be able to sort out some of the simpler knots according to their functional components, the results will not tell us how the components are linked, nor will it be able to describe many of the Xaraxian tangles.  But perhaps as we start to sort out the knots into groups - all those with two black, two white and two green, etc. maybe some bright sparks in our numbers will start to identify basic patterns of performance (security, strength, identifiability, untie-ability...) - lets see.

It is said that every journey starts with the first step - I have been stumbling over 'First steps' for years now - I wonder what this one will bring?  Perhaps nothing more than another bruised knee - but here we go.

Derek
« Last Edit: January 08, 2016, 01:40:07 PM by SS369 »

Sweeney

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 975
Re: Classification.
« Reply #12 on: January 09, 2016, 11:05:50 AM »

So, in line with my previous post, I am producing a set of descriptions and images of coloured knotting Lego bricks.  Then we can all have a play and see how far the idea takes us.  There will be problems, knot structures that just don't fit the simplistic Lego brick concept.


My grandson has built a large number of (very expensive!) Lego models and few Lego "bricks" are used on these. A typcal large model has a huge number of specialised components often unique to that particular kit - have a look at this one for example:http://shop.lego.com/en-GB/Mobile-Crane-MK-II-42009 - there is no limit to the number of components in the sort of catalogue you have in mind.

Sweeney

agent_smith

  • Sr. Member
  • *****
  • Posts: 1391
Re: Classification.
« Reply #13 on: January 10, 2016, 11:02:17 PM »
I hope that this thread doesn't lose steam.

I agree in principle with what Derek is proposing.

I can also sense resistance from some...

Quote
So, in line with my previous post, I am producing a set of descriptions and images of coloured knotting Lego bricks.  Then we can all have a play and see how far the idea takes us.  There will be problems, knot structures that just don't fit the simplistic Lego brick concept.  We will have to either put them in the 'Other' Group (AKA 'too hard to get my brain around') or try to identify and describe new components we come across.

The biggest problem I foresee, is that this is only a partial solution, because although we will be able to sort out some of the simpler knots according to their functional components, the results will not tell us how the components are linked, nor will it be able to describe many of the Xaraxian tangles.  But perhaps as we start to sort out the knots into groups - all those with two black, two white and two green, etc. maybe some bright sparks in our numbers will start to identify basic patterns of performance (security, strength, identifiability, untie-ability...) - lets see.


Am most interested to see how this develops... I would add that the theory must be able to capture all types of knot structures - include the 'Xaraxian tangles' (otherwise it would be invalid).

DerekSmith

  • IGKT Member
  • Sr. Member
  • *****
  • Posts: 1555
  • Knot Botherer
    • ALbion Alliance
Re: Classification.
« Reply #14 on: January 11, 2016, 02:36:37 PM »

Am most interested to see how this develops... I would add that the theory must be able to capture all types of knot structures - include the 'Xaraxian tangles' (otherwise it would be invalid).

Well, it is a Method rather than a Theory.  A method aimed at yielding understanding and classification, but it won't be perfect, and it certainly won't be all encompassing.  We might hope that in the longer term we will develop our understanding to encompass the more complex structures, but initially we need to cut our teeth and explore the viability of the method on the simplest of knots.

Derek