Saturday, 10 April 2010

Notes on Three Threads

I slight change from my usual collection of low-quality jokes and number. Today Gary Gillman discusses Three Threads.

We'll return to my junk tomorrow.

Notes on Three Threads and Numerical Variations

April 8, 2010

Gary M. Gillman, Toronto, Ontario

Below are some sources, both distant and more recent, which indicate that "three threads", the supposed antecedent to London Porter, was just one of numerous thread variations. These sources indicate that there existed a single or one thread drink (probably small beer or ale), and beers known as two threads, three threads, four threads and six threads.

Here is an 1896 article from the New Zealand press which I later deduced comes from a 1727 (not 1798) book called "A Dissertation upon Drunkenness". This is a work of anti-drink literature, possibly written by Bernard Mandeville, a social thinker and physician.

1896 article from New Zealand

In this passage, porter is mentioned and is termed "plain" and "humble". I found interesting, first, that the term "plain porter", used into the 1900's in Ireland, has such a long lineage. In addition, amongst a host of fascinating beer names, both three threads and six threads are mentioned. These thread drinks are stated moreover to be "compositions of sundry liquors". Clearly, different beers were combined in 1727 under these denominations, confirmation of what later writers suggested in this respect.

In the following news article printed in 1904 also in New Zealand, reference is made to three threads and four threads, amongst yet further picturesque beer terms. These derive (the second quotation in the article) from a 1698 book by William King called "A Journey to London in the Year 1698" ("Monsieur" Sorbiere was an assumed name).

1904 article from New Zealand

Some of the beer references in "A Journey to London" were picked up in a couple of 1800's texts. See e.g., James Hogg's "London Society: An Illustrated Magazine of Light and Amusing Literature", 1866, at p. 334. William Black, in his "Practical Treatise on Brewing" of the 1840's, referred to drinkers in the 1700's ordering "one, two, or three threads, as they termed it". Black ascribed the threads respectively to pale, brown and amber beers, which I believe was an over-general conclusion.

These sources have led to me doubt that thread is a corruption of third, since for example a six threads drink would mean double measure or two quarts, and a two threads drink would mean less than a pot of beer, which does not make sense to me.

Recently, in the course of discussing aspects of porter history with Dr. James Sumner, an authority on the subject and Lecturer in History of Technology, Centre for the History of Science, Technology and Medicine, University of Manchester, Dr. Sumner kindly drew my attention to the following essay by E. (Edward) Denneston from 1713:

Denneston essay, 1713

Dr. Sumner pointed out that the essay is cited in Peter Clark's 1983 study, "The English Alehouse: A Social History, 1200-1830". In his essay, Denneston, an official concerned with the collection of excise, recounted his visit to an alehouse in Goodman's Field (part of current Aldgate area of London). The alehouse was called Fortune of War and had a sign which offered " two thrids, three thrids, four thrids, and six thrids".

Denneston, by information obtained from the publican, concluded that the numbers in each thrid or thread type related to price. Thus, a pot of two threads was sold for two pence, three threads for three pence, and so forth. Moreover, the price asked was in proportion to strength. Also, each thrid variation was mixed from two types only, strong beer and small (and apparently beer, not ale of any kind, although Denneston does not recount any further details in this regard).

Dr. Clark appears to have considered that "thrid" was a variation of "thread", which I would agree with. In referencing the Denneston essay, Dr. Clark was adverting to the contemporary practice of mixing strong and weak beers to avoid the full weight of taxation on beer and ale. The essay is a well-reasoned plea to establish a more effective system of beer taxation. At the time, and as is well-known amongst English beer history buffs, a tax was established for strong beer and a much lower one for small beer. In turn, the categories of strong and small were determined by the selling price of these beers. Brewers, and publicans who brewed for sale, had an incentive to mix small beer - sometimes unfermented wort was used - with strong to sell the blend as strong but avoid paying full tax. For example, if they brewed two barrels of weak and one of very strong beer, and paid the correlative rates but blended and sold the totality as strong, tax was evidently underpaid by a heavy margin. Denneston drew attention to this and numerous other underhand practices of the day.

The Denneston piece does not refer to the strong beer being "stale" or to any balance of palates (flavours) sought in the mixing. The object, as we can glean it from this essay of 1713, was to offer beers of a certain strength, calculated in relation to price. Now, the publican of the Fortune of War was practicing a deception. Could the threads - four types - have corresponded in honest pubs to different types of beer? To all "sorts" of drink, as the publican claimed when Denneston walked in? If so, this would suggest that the threads generally were different beers and were mixed both to attain a certain palate and perhaps a certain alcohol level. Thus, six threads would have been six beers mixed in the pot. However, I think this was unlikely. This is not to say, as I state below, that flavour may not have played a role in drinkers' preferences amongst the thread group.

In light of the Denneston essay, I believe therefore it is not likely that thread was a corruption of third and if anything, the reverse is true. However, it is still possible that the traditional assumption that thread is a variant spelling of third is correct. Perhaps three thirds was viewed circa-1700 as a full measure of alcohol (or perhaps I should say an entire measure). And therefore, six thirds can be viewed as double measure, not in volume, but in strength. Two threads would have been one-third less alcohol than full measure, and so on. I know too that the term third in its earlier history was sometimes spelled thrid. This interpretation is possible but I incline against it.

A number of suggestions have been made in beer historical literature concerning the derivation of a non-third thread, which I won't canvass here. An explanation not previously advanced (to the best of my knowledge), might run as follows however.

The following statute from the 1690's, intended to prohibit the use of salt in brewing, employs the term "brewing-loom" to refer to an apparatus of brewing equipment:

1690's Statute Referring to Brewing-loom

Loom at the time meant, not just a loom for weaving, but any industrial or artisan utensil, see e.g., Bailey's mid-1700's etymology dictionary. One can conceive that an informal usage developed among alehouse devotees to call a beer a "thread", metaphorically by reference to the general meaning of loom since even at the time its primary meaning was the one we associate with the term today, for weaving. While this etymology may sound (and be!) unlikely, in fact a number of terms for beer in the early 1700's seem to pertain to textiles or weaving terminology. For example, stitch, meaning strong brown ale, was and is a term used by tailors. Some cant dictionaries rendered the full term as stitch-back. A devoted tavern-goer was sometimes called a malt-worm, and of course worms of the animal kingdom played a role in the process of making silk threads. There is at least one reference in 1700's writings to spinning a thread to mean drawing a pot of beer.

What may be more intriguing is porter - and indeed the term beer itself - were also technical terms used in weaving to mean a device which held multiple threads:

History of cotton weaving (1895) Richard Marsden

The above 1800's work indicates that porter used in this sense was a term familiar in Scotland, and beer was the term used in England, but one can speculate that both terms were also known in London in this sense. Perhaps the idea to call different forms of beer by reference to terms familiar in weaving and tailoring started with the realisation that beer itself had a double meaning. Could this beer-naming system have developed as a kind of elaborate in-joke, perhaps reflective of East London humour at the time? Cockney rhyming slang is not the same thing but one can envision a rough analogy. "Portering" threads was a way of processing or combining them (in weaving again, per the source above). Pamela Sambrook, in her fascinating "Country House Brewing in England, 1500 -1900" (1996), reports a circa-1700 usage of the term "portering fatt" in one country house inventory, a vat which apparently was used in brewing operations.

Sambrook History of Country House Brewing

I find the term portering somewhat vague, but given that in 1600's weaving terminology it was a combining or processing of threads, I wonder if the term was borrowed by brewers to describe a combining of beer "threads" - product of the brewing-looms - to attain a specific quality. The combining could have occurred simply by mixing different beers and sending them out, or possibly by entire gyle brewing. And if so, did that in turn spark the development of a noun, porter, to describe such mixed beers and the later entire version?.

I wish to emphasise that the connection to weaving and tailoring is speculative, but one I find interesting.

Returning to the idea of multiple thread beers, with the benefit of the Denneston essay, it seems unlikely again that the thread beers contained more than two beers. They might have on occasion, but I don't think it was essential to their character, speaking that is of three threads and higher. A mid-1700's definition of three threads confirms this. Bailey's etymology dictionary from 1776 gives the following definition of three threads:

Bailey's Dictionary

Under this definition, three threads is common ale and one of two other beers. This definition can be read as palate-related and strength-related or just the latter, but strength regulation seems key to this definition.

In Ian Hornsey's "A History of Beer and Brewing" (2003), the author cites a statement by one Wagner from 1924 which refers to a "two-thirds" brew as a mixture of strong ale and weak beer. Wagner goes on to suggest that entire was devised by Ralph Harwood in 1722 to replace two-thirds and impliedly three threads (or thirds) which was two-thirds but with pale ale added. This account from Wagner is more in line with John Feltham's account from "A Picture of London" (early 1800's) and Obadiah Poundage's famous 1760 letter. I mention it here simply to show that two-thirds - the beer of that name - cannot have meant 2/3rds of a pot of beer. I believe it meant not just a half and half but a two pence the pot beer, one with a relatively low alcohol content.

Of course, it is possible some persons who sold thread beers did mean for each to offer a specific taste, not just (or even at all) a specific strength. It is also possible though that by around 1760, people forgot what the thread combinations originally meant. Even after 1760, however, there is some indication that people remembered that the threads signified what they do in Denneston's essay. This brief note from 1823, headed "Calvert's Butt", suggests this:

1823 note on porter development

(His reference to Calvert as originator of porter in contrast to Ralph Harwood puts the claims of Harwood's invention into further doubt I think. The account states it is to be continued, but I could not locate a sequel).

I think one might infer from all this that the thread drinks were mixed by the brewer or brewer-publican primarily to avoid the full tax burden on beer; that after Denneston wrote (1713), the authorities were more alive to sharp practice, and therefore entire beers may have appeared as a lawful, safe substitute. In Martyn Cornell's discussion in his books and other writings of the "malt-worm" pub guides written between 1716 and 1720, he indicates the term entire was used in one of the guides to describe a beer offered. I believe this is the earliest use of the term known in a consumer context. "Threads three" (a poetic rendition of three threads) is also mentioned in one of those guides. Perhaps for a time both versions were available until entire effaced the threads. From my own searches, all online conducted and non-systematically, there seems little mention of the thread drinks, except historically, after about 1730.

I believe an entire porter was made, as Martyn Cornell has argued, by combining multiple mashes. For efficiency and other reasons, I think the result would likely have produced an average-gravity beer (entire gyle small beer apart), which would have copied the strength range of three threads. And this would have been around 5% ABV judging by the latter's place in the thread range. We know from Poundage that porter was first sold for three pence the pot - the same price as three threads according to Denneston, so the replacement idea fits in here. Why though would the thread beers not just have died away due to tax man pressure, to be replaced by the weak and strong beers or ales which preceded them? Possibly the latter did not offer in palate perhaps, but especially in strength, the desired mean, the "sweet spot" that three threads hit. I don't know what small beer was in alcohol then but even if it was 2% and strong beer was 8%, three threads might have hit a norm of 5-6%. I would think the thread range might have been something like: 2% for single thread; 3% for two threads; 4-5% for three threads; 5-6% for four threads; and 8-10% for six threads.

So this way of looking at it looks mainly at the alcohol level of the threads and that entire butt substituted for the most popular version, the mean.

Alternatively, perhaps porter and entire had nothing to do with three threads. The thread system may have been a tax dodge plain and simple. I'll give you strong and stronger at a good price, never mind what it tastes like (may have been the idea). In contrast, porter and entire may have been improved, extra-hopped brown beer full stop, as some beer historians have argued persuasively. There may have been the intent, as some also have argued, to minimise the tax placed on malt in the first years of the 1700's by using less of it and more hops than brown ale used, but that is different from the subterfuges to which Denneston alluded. Perhaps therefore Poundage and later John Feltham confused the thread beer phase and rationale with early porter development. This would have been understandable since porter, like the thread beers, was a mixture, either literally through the continued practice of mixing mild and stale beer, or through long aging of well-hopped beer which resulted in a narrowing of the "extremes" Poundage mentioned.

And yet, it may have been a case of a bit of both.

Where does this leave the apparent early role of high-dried brown malt? I am not sure."Porter's liquors", also mentioned in one of the malt-worm guides cited by Martyn Cornell, may have been characterised by such malt - an inexpensive malt which ended up creating, or helping to create, a new style of beer. If high-dried brown malt was the key to porter development, the threads, as explained in Denneston's essay, seem even less relevant. However, I incline to think that brown malt was always present in London's basic beer system, the beers the porters and other manual workers famously favoured. Graham Wheeler has argued this in comments made on Ron Pattinson's beer blog and I believe he must be right. True, perhaps an extra-dried version emerged in the early 1700's, but surely there would have been variations all along in the fermentable extract and diastatic potential of the brown malt used in London for the basic ales and beers.

Some kind of brown malt was used for example in the thick, cloudy, strong brown beer Michael Combrune found wanting. I believe other factors account principally for porter's development, which is not to say that brown malt well-dried did not end by being a noted key-note flavour in London porter (of course it did).

A last thought. If mixing strong and weak beer was disallowed to brewers, and publicans who brewed, what explains the suggestions in the Feltham and Poundage accounts of porter origin that the pre-porter drinks were mixed? One answer may be that the law, at least at one point, exempted mixing small amounts. Tracing the history of beer taxes and the mixing prohibition is a complex task, one I have not engaged in or propose to at this time. However, I did find this mid-1700's statute, which seems to permit retailers of beer to mix in quantities of three gallons or less:

1700's Statute on Beer Mixing

Thus, if the authorities cracked down periodically on beer mixing, maybe Parliament at some point in the 1700's changed the law to let pubs mix beers for drinkers at their request.

End Note: I learned of Dr. James Sumner and his work through Martyn Cornell's well-known beer blog, The Zythophile - Beer Now and Then. I want to thank Dr. Sumner for taking the time from a busy schedule to communicate with me. Martyn Cornell's impressive writing on beer history, as well as that of Ron Pattinson on his blog, Shut Up About Barclay Perkins, have been indispensable to understanding numerous key issues in porter history as has Dr. Sumner's compelling article, "Status, scale and secret ingredients: the retrospective invention of London porter", published in 2008 in the journal History and Technology. Also needing mention are the brewing history (or related) writings of Peter Clark, H.S. Corran, Terry Foster, Ian Hornsey, David Hughes, Michael Jackson, Peter Mathias, Roger Protz and R.G. Wilson amongst others. Thanks finally to Ron Pattinson for kindly agreeing to place these "Notes on Three Threads" on his blog.


Rob Sterowski said...

Ah, I have only skimmed this post as yet, but it's already been helpful in understanding how the tax scam of mixing strong and small beer worked, something that was puzzling me reading another text.

Bill said...

"1698 book by William King"

I don't remember writing that book.

Bill King aka 'First Stater'

Mark Oregonensis said...

That is certainly the first I have read about a possible connection with weaving terminology. Thank you for this. I am brewing Ron's, Kristen's and Barclay Perkins' 1848 TT tomorrow, and this will serve as some fine inspiration.

Gary Gillman said...

To assist readers further, here is the Sorbiere/King book from 1698:


Gary Gillman said...

Due to a comment privately received, it was suggested that the early porter, based on late 1700's published analyses, might have been 6.5-7% ABV. My suggestion that 3 threads/first porter was 4-5% ABV would be low in this light.

My logic in selecting that range was twofold: I took 10% as the norm for 6 threads - thinking in practice it couldn't go much higher - and halved it. Also in my mind was the possible lower alcohol yields of circa-1722, when porter by many accounts was not "improved" until much later in the century through better malts and longer standing.

However, of course 6 threads might have been 12% ABV and anyway the strengths may not have been proportional over the thread range as I had assumed. Early porter may indeed have been around 6% ABV, and perhaps brown stout then was 8%.

This is an area where one must speculate... Probably too the early entires did differ in their ABV.


Graham Wheeler said...

This three threads thing is worthy of further discussion. It is a shame that this is about to drop off the radar.

Although I do not necessarily agree with some of the conclusions that Gary put forward, he has pointed out, quite positively, that we have not got a bloody clue as to what three-threads really was.

This three-threads thing is fundamental to understanding the history of beer. I have never believed that three-threads was the forerunner of porter, for reasons too complicated and too controversial to go into here, but nevertheless it has its place in history. Writers and, so called, historians have taken this three-threads thing at face value without questioning it, I have done the same, but Gary has shown that there is a lot more to it than the folklore supposes.

Getting to grips as to what three threads, or any number of threads, really was, is fundamental.

Any number of original researcher on here should have said to themselves: "Hang on, this is quite important" - but it has been allowed to drop. Shame!

This is a good case for Patto setting up a proper discussion forum so that important things can be thrashed out, without dropping off the radar prematurely.

I'd even set up a forum for him, but methinks he's probably an IT professional anyway, and can do it for himself. I do not suppose that there are many ex-pat Brit daffodil growers or Brit water engineers in Holland.

Gary Gillman:
Well Gary, may I thank you personally for the various documents you have pointed us to. How you find them is a mystery to me; I never have the same sort of success whenever I search. I can only assume that you have witchcraft on your side.

One way or another, some of those documents (those that I can get at) have improved or modified my knowledge for the better (well, actually, considerably).

Thank you.

Gary Gillman said...

Thanks Graham for your kind words.
I did a lot of searching in Google Books and some other publicly available databases. If you do that using different key-word combinations a lot can come up. The all-important Denneston essay was something I did not find: James Sumner told me about it as I recounted earlier.

The weaving terminology part came from just thinking about thread and stitch-back (and malt-worm, which I first read about in Martyn Cornell's books) and then doing different kinds of search combinations. The brewing-loom term came up by accident: I was looking for information on salt in beer.

I really just spent a lot of time searching and thinking, that's all.

I know many of my ideas are open to question - and I stated the weaving connection is still speculative - but I wanted just to put out my thoughts and open up hopefully some discussion.


Gary Gillman said...

And I must say again: the current crop of beer and brewing industry historians are doing superb work. One can't even think of making progress in this area without absorbing and understanding what they have done.

My original interest in beer history was for one reason only, one that I think still comes through in many of the comments I've made on this site and others: I am interested to know what porter and other beer tasted like in the eras under discussion. That was and is the spur of most of my thinking.



Gary Gillman said...

Although now in "Older Posts", I'd ask you, Ron, to please insert this last comment from me, to mention that I just noticed that the Denneston essay is also included in the bibliography of Peter Mathias' landmark book, "The Brewing Industry in England, 1700-1830" (published I believe in 1959). Denneston in the bibliography is spelled "Denniston".

I have not been able as yet to check whether Dr. Mathias discusses thread variations (i.e., those other than three) in the text of his book, to the best of my recollection he does not. Peter Clark quoted in his history of the English alehouse book the references to "...two thrids, three thrids, four thrids and six thrids" from the Denneston essay, rendering "thrid" as "thread" in each case (from what I can tell from reading partial views of Dr. Clark's book online).

The full title of the essay by E. Denneston (this spelling appears on the title page) is, "A Scheme For Advancing and Improving The Ancient and Noble Revenue of Excise Upon Beer, Ale and Other Branches".

Perhaps future researchers and beer writers will address why five thrids/threads was not mentioned by Denneston. Perhaps the Fortune of War was simply out.