Submitted by Frank Ellis, Fred Kincaid, Fred O’Leary and Joe Paradi

Article 3 of a series
The SCReW Decision Tree

The first two articles of this series on SCReW, the Standard Corkscrew Reference Work (PART 1 and PART 2), covered the background to the project and the general structure of the database. They also highlighted the benefits to all corkscrew collectors of a massive repository of information that can be manipulated by anyone for his or her own purposes.

The core of the database is the SCReWcode that is applied to each corkscrew whether it be a humble straight pull or a fantastic elaborate patent. The SCReWcode allows rational ordering and record keeping within the database and a means of searching for individual records. Use of key words is another way of searching.

Within SCReW, all corkscrews and associated collectibles are organized into 13 Classes, each designated by an upper case letter from the name of the Class itself (except one). Although the Classes were covered in Article 1 it is worth looking at them again. Ten of the Classes are for corkscrews, one for Non-worm Cork Pullers, one for Champagne Taps and Tools and one for Accessories.






Collectibles and “go-withs” which are not actually corkscrews but which we all seem to acquire


Bar screws

Mechanical corkscrews that are designed to be fixed to a wall or a bar



Corkscrews with a tool or a design having one or more secondary functions



Corkscrews designed to ease the adhesion of the cork before pulling it out with a direct pull



Corkscrews in which the whole item, or a significant part of it, represents a tangible object



Knives (folding or fixed blade) with a folding corkscrew



Corkscrews that have a mechanical action through the use of a simple lever and fulcrum



Corkscrews which have a mechanical feature using a thread, ratchet or rack and pinion action


Non-worm extractor

Cork removers that do not have helical worms


Pocket and Protected

Corkscrews in which the worm is covered or otherwise protected for safety or travel


Self pullers and partial pullers

Corkscrews based on the principle that continuous turning of the worm draws the cork up the helix


T-screws and other straight pulls

Corkscrew with no mechanical features and no other tools


Champagne and soda tools

Tools for opening or tapping sparkling drinks bottles

It all looks pretty straightforward but the key to making the system work is to get the corkscrew into the correct SCReW Class. That is not always as easy as it sounds. Some corkscrews have very obvious homes; a Thomason goes into M for Mechanical; a Rotary Eclipse is B for Bar Screw; a Williamson with bell is S – a Self Puller; and we can all recognize a simple straight pull that is coded T – or can we? Read on.

Just a moment’s thought however will highlight some potential areas of ambiguity. Where does a HOOTCH-OWL fit? Is it L since it is a double Lever? Is it F since it is Figural? Is it C to account for it being a Combination tool with cap lifters and bottle top grips, among other tools? Where would you put a straight pull with a spike on the handle?

As discussed in Article 2, the decision should be relatively intuitive but even the four of us had vigorously differing opinions on some items. SCReW uses two basic principles to code corkscrews:

1. Any aspect of the corkscrew that provides a mechanical advantage to removing the cork takes precedence over those which do not, e.g. the lever action of a HOOTCH-OWL takes precedence over Figural and Combination tool considerations – it is coded as L.

2. The “essence” of the corkscrew will normally be a deciding factor.

The sheer inventiveness of corkscrew designers has posed many problems for those who aspire to neat and tidy classification – even with these two principles at work. Rules have been devised to determine where “difficult-to-decide” examples are placed. But rules can become unwieldy for everyday use. From necessity was born the SCReW Decision Tree.

The SCReW Decision Tree is easy to use and will put at least 99% of all known corkscrews into the “right” Class where they can then be number coded. “Right” is a judgmental term, of course, and some collectors will disagree with the decisions on marginal items. Accept it. There will always be disagreements. The aim of the Decision Tree is to make life easy; to follow step-by-step questions with “Yes” and “No” answers; to keep the wording brief yet the meaning clear. From extensive testing it has gone through an evolutionary process to arrive in its present form. And it works!

The guiding principle of “mechanical advantage always wins” is broken only once. Many Syroco figures have a bell and hence are self pullers in operation. However they are a niche collecting area and their “essence” is that of Figural (see Bull, page 249). The mechanical purist may not feel comfortable with this – but there it is. The Syroco family SCReWs together!

Try some examples on the following pages and see how the SCReW Decision Tree works. The HOOTCH-OWL is clearly not an accessory, bar screw, non-worm extractor nor champagne tap. But at the next question it is a lever and so is coded as L.

A Syroco Waiter is not a lever, nor has it got cogs or threads so the Decision Tree directs you to the next question. It is a self puller and it has got a Syroco “handle” – it is therefore coded under F. Think about some more examples and try out the SCReW Decision Tree. But don’t forget – always start at the top or you may fall into traps!

So go and try the SCReW Decision Tree

Did it work for you? Hopefully you found that most things end up exactly where you would expect them. You only really need the Decision Tree for the “border-line” cases. Did you discover any? Were there any quirks?

Not quirks – but there are some things I would like to know more about. Why is the third question worded “Has it got cogs and/or threads to help pull out the cork or insert the worm?” I thought that Mechanical referred to mechanical advantage for removing the cork.

Within SCReW, Mechanical is the word that describes a Class of corkscrews that have the mechanical properties of threads, gears, racks etc. It does not imply that all corkscrews within M provide a mechanical advantage for removing the cork – although the majority of them do.

For example, the “essence” effect diverts from engineering erudition in the case of the registered corkscrews of Robert Jones and the similar Thomason IV (Wallis page 17). The purist will regard these as Easers since the user has to use brute force to pull out the cork. However we believe that most collectors will see them as Mechanicals due to the threaded shank – they therefore end up in M. The same argument applies to ratchet handles (Newton and Dixon; Wallis page 36; O’Leary page 36) and direct pressure insertion corkscrews (Qvarnström and Hull; Wallis pages 174, 354). The key message is that the SCReW Class names are just names – nothing more. They are a means to an end (finding records), not an end in itself.

It looks like all double helix corkscrews end up in the Easer Class. Does this include roundlets, folding ones, bows and even the peg and worm?

Yes. The essential part is the double helix – and that functions as an Easer.

The Easer line looks complicated. What’s all this about sheaths and “significant” tools?

Easers are far from a neat and tidy classification. Much ambiguity could exist without bringing in the Decision Tree. Remember that anything with a button is an Easer – it will give the cork a twist once it is screwed home. But some of these buttons, especially on wooden sheathed Cloughs are probably only there to stop the worm from pushing through the bottom of the sheath or the cork itself. They were designed as pocket corkscrews and “essence” directs them to P. There is an apparent contradiction in the otherwise “perfect” system here. Millions of these wire Cloughs with buttons – particularly the smaller versions – were made without sheaths as disposable cork rings for vials and medicine bottles, to be kept in the cork for recorking. Moreover, many sheaths have broken or separated from their original corkscrew partner. The result is that the corkscrew world has become flooded with almost identical wire corkscrews, some with a sheath and some without a sheath.

So how was that mess resolved?

In this one case the same corkscrew will have more than one SCReWcode – one without the sheath under E and one with the sheath under P; and it may also turn up in C.

Why C?

The situation is even more complex when you think about all of those wooden handles with tools on them – especially with a Clough wire corkscrew with button stuck in the end (see O’Leary page 114).

Yes, I’d wondered about them. They’ve got to be combination tools.

Dead right – but only if the tool is “significant”. If it is just a bent piece of wire that acts as a cap lifter then it is ignored as another tool and the corkscrew ends up in P for Pocket/Protected.

But this all sounds so complicated.

Agreed. That’s why the simple step-by-step Decision Tree was designed. It takes the worry out of the complexities. It also ensures that every collector goes to the same place for a border-line item.

Knives must be easy and obvious. Why have they got a secondary question?

Well, yes … but … where is the cross-over between a pocket knife with folding blades and other tools, and a cap lifter that happens to have a folding blade(s) and is clearly a Combination tool? That’s why there are the carefully selected words “Has it got a significant tool extending beyond the main body which is not hinged?” This wording puts Swiss Army knives within K but sends pretty obvious combination tools to C. Have a look at the following examples and see if you can see the point.

I see that there are more complexities about folding tools in the “legs and arms” line.

True – folding corkscrews with tools; combination tools; and folding Figurals, both with and without tools, are a challenge. But if you follow the Decision Tree it all make sense. Just consider the following three items. Both the ladies legs and the “Tip Top” (left & center) are coded P as the “legs/arms” protect the worm. The cap lifter on the right is similar but has a significant non-folding tool, which directs it to Combinations. These are difficult border-line cases.

Hey, I’m getting the hang of this. The Decision Tree pulls out roundlets and walking sticks before Figurals so that I don’t have to anguish over whether a bottle shaped roundlet is a Figural or a Pocket. And the same goes for that cane with the Howard Luterman handle (Bull page 40) – he’s an old stick in P and not a Figural beauty.

You’ve got it – but be careful how you put that to Howard.

I guess Figurals show up all over the place but are pretty straight forward once you get to that line. Lots of figural double levers have been pulled out into L together with mermaid waiter’s friends before you get to this point; the ladies legs and the bullet roundlets have gone into P; so all I’m left with in F are those British brass things.

Not quite. There are lots of Figurals. In fact it is now and will likely remain the largest Class thanks to the whims and fancies of corkscrew producers around the world – particularly the 20th century – ranging from golf balls to erotics with balls; from dogs to hot-dogs; from heads to feet; and from Anri to zebra.

That pretty well covers the alphabet.

You got that right! But you should note one quirk in the Figural group. Corkscrews with sheaths that have silver figure handles are all grouped together under Pocket. That is to bring together like with like. The Dutch “man with cow” handle (de Sanctis & Fantoni, “The Corkscrew”, page 156) and elegant Dutch and English silver sheathed Ts (Watney & Babbidge, page 38) should go together.

So that then takes us down to the Pocket and Protected group. That all looks simple and I can see the sense of keeping all bows together. There is no point putting single worm bows in P and multi-tool bows in Combinations.

That’s “essence” working for you. But don’t forget – size is not relevant to SCReW – well at least in this context of the word!

So what comes next? I see that there is one of these “significant” tool comments again in the sheathed group.

Yes, this is to make sure that sheathed things, usually picnics, with a small cap lifter hook on the head ends up in Pocket, but the bigger tools that happen to have a corkscrew stuck into the handle are placed in Combinations.

And combinations? – Are there any corkscrews left to go in there?

Hundreds – a ton of can openers, cap lifters, spoons and spikes – but for reasons of economy of space SCReW does not consider a brush on a handle as a “tool”. Brush handles fit in alongside other simple straight pulls in Class T.

So the trick is to ignore the complexities and just follow the branches of the Tree to reach a simple conclusion. Are there any other problems with the SCReWDecision Tree that I should know about?

Only one. It should not be used on public transport. One of us was using it on a train and the middle aged American lady in the next seat became more and more intrigued. Finally she asked very loudly “a screw decision tree – gee, what kind o’ business are you in?”

References to books:
British corkscrew Patents from 1795, Fletcher Wallis, 1997
Corkscrews 1000 Patented Ways to Open a Bottle, Fred O’Leary, 1996
The Corkscrew, Paolo de Sanctis & Maurizio Fantoni, 1990
The Ultimate Corkscrew Book, Donald A. Bull, 1999