William Caslon, Typefounder


# 1
# 2
# 3
Caslon, No. 337
# 4
# 5


Caslon Oldstyle 337


'The Oldest Living Typeface'
William Caslon


FOLLOW UP; This conversation continued in an internet discussion. Thankfully John Hudson re-positioned himself with a suggestion that does away with the ink spread theory, some others have not.

This text is waiting for updating.

CASLON 337 was the most ambitious production of a digital face up to the time of its launch. Lanston's digital release expresses in minutiae Lanston's punch cutting replication of Caslon Oldstyle in 1915.

Caslon 337 was very well received by the type community. However, in this writing I am going to address an issue in which John Hudson, Tiro Typeworks, mis-spoke about ink spread and the bolding of type. Type designers that engage themselves in type revivals make much ado about ink spread on hand made papers.

And John is so well respected in the type community I fear following generations will parrot the error into fact. So I have taken this opportunity to set the matter straight. Trusting John will understand this is not a personal attack on him.

And, trusting John's pension for correctness, he will thank me for making note of those errors.

'Although the forms of Adobe Caslon have been somewhat 'cleaned up', compared to the delightfully eccentric 18th century forms of the original, it does possess a better weight than, for example, the Lanston Type Co.'s Caslon 337 (which is much more historically accurate). The Lanston fonts were digitised from smoke proofs of metal patterns, and as such carry none of the weight associated with ink spread. They are, however, ideal for letterpress printing from polymer plates. '

John Hudson


I AM AT GREAT VARIANCE with John's observations about weight .

Certainly, I for one, do not prefer Adobe's Caslon over Lanston's Caslon 337 in text sizes. Also I do not subscribe to any puerile concept that William Caslon intended his face to become bolder by inferior press work.

But I do agree that Caslon 337 is light and difficult to read if you choose six pt. Caslon 337 for your setting.

But that is because you are not setting six point, you are setting 11 on 12 point linear scaled to 5 on 6 pt.

Caslon 337 six pt.'s x height is equal to less than most faces of 5 pt. Type that small is excellent for labelling dangerous products or for redistributing the wealth of a centenarian. There should be a law against type that small. In some countries there is.

But can you imagine the remarks had we released five on six point, linear scaled to twelve point.

Anyway, if John had stated he preferred the weight of Adobe Caslon in six point over Lanston's Caslon 337 in six point I may have even agreed with him. What joy is there to be right for the wrong reason?

If you want a closer to source Caslon than Lanston's Caslon Oldstyle 337 for setting text you'd best start shopping for hand-set foundry type. Lanston's revival of Caslon employed uncommon type revival techniques. An account of the project was duly noted and published in my typographical journal, 'The Fount'.

It was of great interest to me that another foundry digitised the various sizes of the original Caslon. Disappointingly the gentleman reproduced Caslon with ink globes and all. Hopefully his work was not derivative of the dubious ink spread theory that was spread around like peanut butter in internet discussions groups.

He would have saved himself a lot of trouble if he were aware that Caslon Antique was readily available.

I suspect my complaint would be minimized when this version of Caslon Oldstyle is output at their appropriate point sizes. Some typesetting houses may find the settting somewhat confusing.


WILLIAM CASLON'S cutting has some very interesting qualities. Different designs for different point sizes. This includes intuitive optical scaling but also contains other ingredients of early type cutting practices. I suggest however such an illustration could be achieved simply by using the original hot metal type. Hasn't anyone heard of letterpress?

His work is interesting in illustrating a bygone type design philosophy. If he had not preserved, 'as precious' the 'ink smear' I would have held his work in the highest of regard.

Sometimes there may be use for his cutting. The movie industry for instance. Or for illustrating a type history book. A bookbinder doing repairs. An example of optical scaling. It's sounding better all the time. I could actually use it for this article.

I would like to offer it in the Lanston Library. It is interesting, worth looking at. It is self limiting though.

Here is a look at our competitors Caslon.

Justin Howes, Founder's Caslon


JUSTIN HOWES has been quoted as saying that he considered re-cutting his typeface for the letterpress 'polymer printing process'. Hopefully this would make his new improved Caslon much like ours, which is correct. Probably his cutting would have a tighter fit at 11 on 12 pt. than Lanston's.

As many know, we serviced our hot metal customers long after the others stopped. The reasons for a looser fit is a lengthy technical discussion on its own. I will not include this discussion, or duplicate it here. It is recorded in 'The Fount'

Anyway if Andrew were to pursue this cutting he would wish to have at his disposal the customized optical tools we used at Lanston. Believe me, it was quite the contraption, very long and with no other purpose than to inspect type or matrices at extreme enlargement. All this without measurable evidence of parallax.

In any event his admission that his types are not right for the polymer printing process bodes poorly for his cutting of Caslon Oldstyle.

Has it not already been indicated that Caslon Oldstyle 337 looks correct to John Hudson's eye when printed by the polymer letterpress process?

Note: For the sake of argument we are discussing Caslon 337 in the intended text size.

And has it not already been 'previously agreed', Caslon Oldstyle is ideally suited for letterpress printing?

Could I be brave enough to suggest that letterpress is superior to computer technology.

Apparently I am.

Note: Point Sizes were not standardized and therefore these are approximations of point sizes. This is a discussion entirely on its own. I think we have, for the moment, enough to deal with.

If you are going to print a book in six point you could consider Adobe's Caslon. Something tells me that's not often going to happen. My work with Caslon 337 was primarily for printing books, the hot metal version was 'optical scaled' for six point. Caslon 437 was used for large display sizes. Remember, I am speaking of letterpress printing.

I don't want to ruin my reputation as a grump but Adobe Caslon is a pristine revival that preserves much of the feeling of the original. Carol Twombly has earned my greatest respect for her work. As a re-drawing, Adobe Caslon is singular in capturing the feeling. I include comparisons to even most hot metal 're-cuttings'.

Puritanically speaking only Caslon could have been aware of his decision making processes while cutting his own faces.

Type interpretation falls towards readership rather than authorship. The movie is less likely better than the book.

Optical scaling for Caslon was visual inspection, a craftsmanship instinct to make things look right. In other words subliminal and natural. His designs were not hindered by intended mathematical correctness. 'Hinting type faces' at this juncture of type cutting was biological. The traditional proofing medium for hand cut type was smoked proofs. Smoked proofs are crisp, clean and razor sharp.

Alright, don't smear them.

Did you ever wonder how punchcutters transferred their designs accurately onto the ends of teeny weeny bars of steel? Something to contemplate for sure. I suggest that some type designers just started cutting, some may not have even had drawings. Much like I design, in the stick. I am sure each found methods that suited their temperament. But whatever method was used the artist would need to exercise impossible deftness to produce an accurate same scale drawing of even 18 pt. without considering 12 and 6 pt. of the intended type size. (Not to mention that the punch is a flipped image of the drawing, there are ways to work around that distraction.)

Easier to cut to scale than to draw.

One method used in the early scaling of type is the tools one is born with. Hand and eye. One would anticipate great mathematical discrepancy from drawing to product.

But was it a loss or a gain? That would depend somewhat on the craftsmanship.

Later photo transfer methods were experimented with finding a degree of success. Particularly if acid etched.

Do not misunderstand me. Punchcutters made changes throughout the process. But the intended outline was recorded in metal, not the fluidity of ink spread. That was beyond his control unless he was an in-house typefoundry. Well, not even then.

Ink smear in printing was not in the founders control and was not a sought after attribute of the printing process.

Later, mechanical cut typefaces scaled linearly was found to be disappointing. Type faces were not correct at smaller sizes. They were now aware of optical scaling, previously it was the artisans judgment never formulated mathematically. They eye has optical consideration as does any lens. Corrections are required.

This problem was overcome by making different drawings for different pt. sizes. Caslon 337 was manufactured by this tedious method.

Within this frustrating process of creating multiple drawings and multiple brass patterns, the punchcutting department discovered some linear aspects of optical scaling.

From then on the type design department focused most of their attention on the outer extremities of the pt size range. For instance, 12 point and 6 point.

The inner sizes were stepped.

The punch cutting department re-engineered their systems using different followers, cutters and different X, Y, co-ordinates. These were then recorded on punch cutting cards.

A compromise perhaps, but it worked well becoming the modern model for optical scaling in the mechanical typesetting era.

I do not believe there has been a singular improvement in typography, printing or for that matter, type design, that can be attributed to the digital era. Don't get confused by quantity!


I PREDICT A DARK eternity will soon befall the reading public and become embedded in typography forever. Perhaps your suspicions are that I am sliding off from Caslon, but I am not.

My digital preference is without any doubt Lanston's Caslon 337. However my enthusiasm for digital typography has been greatly dampened by a broken miscalculation in the typographical priorities of the computerized age. I am stunned that some key exponents insist on diminishing the importance of optical scaling in typography.

False confidence in the new typographical age of desktop was earned by the ease in which I re-invigorated typographical enthusiasm for properly drawn small caps, swash, both oldstyle and lining figures, not to mention tied characters. Niceties previously dispensed with by the interloping photo typesetting machine manufacturers.

Lanston was first in offering these 'requirements'. They were warmly welcomed by the discriminating typographer. Other manufacturers were pressured into offering these features as 'expert fonts'. We never treated our clients as amateurs. This is called 'corporate culture'.

You will see they also liked our small cap arrangement which is industry standard.

Now I have been blind sided by active industry resistance to 'optical scaling' .

These same industry icons would dismiss quickly any suggestion that scaled caps were acceptable. That drawn small caps should be done away with. That drawn small caps are too much work. Optical scaling is even more important than drawn small caps. I am not merely proposing a typographical 'nicety'. I am proposing a typographical 'necessity'. I am discussing legibility and readability. I hope we are together sharing the concept that text faces are made to be read, not just to bedazzle.

I know it is novel of me to suggest that type faces are meant to be read. But there it is!

Optical scaling 'trumps' drawn small caps. We need them both.

Caslon 337 has given me an opportunity to revive this needed discussion.

They say optical scaling is not linear scaled. If it were linear scaled we would have what we have now, inferior type. But there are linear features within optical scaling.

Optical scaling requires different axis for linear scaling parameters. One axis for ex height, one for weight, one for fit. Perhaps borrowing an auto design feature from Ikarus would be helpful. Type digitizing tools have many more complex operations than this.

But let us review the short history of desktop publishing. American Type Founders offered optical scaled type in 1988. How did it work? Very well thank you!

Rather than write something new why don't they just give Henri a call.

And I ask, If we could do this with a Benton Punchcutter in the hot metal days why has it become so difficult with all this supposed computing power?

But back to the task at hand.

Possibly if I set six pt. text in Lanston's digital Caslon 337, which is actually five on six pt. or more precisely, eleven on twelve point linearly scaled to five on six, I would appreciate Johns observations. But I never set Goudy Hand Tooled below 18 point either. Part of skilled typography is knowing how to pick the right typeface for the job.

However, in spite of this, Caslon 337 was offered in five on six point for hot metal composition. This was still the 11 on 12 pt design but optically scaled to five on six.

In case you don't know it, I am trying to sell you on the merits of optical scaling.

Lanston considered releasing a different point size for the smaller sizes, 437 for the display sizes which would have put an end to John's preference.

Or would it?

Caslon's original cutting was unique in design for each and every point size. This is not what we set out to accomplish, and this is not the basis of John's complaint.

Not offering a unique design for each and every point size is forgivable, some may say preferable, but the inherent absence of optical scaling is not. Lanston has met stiff resistance to the concept of optical scaling in the OpenType forum. John is a primary spokesman in the forum and so I feel if he has objections to weight he is in a position to do something about it.

However his criticism alludes to a topic bandied around the type design community about very legitimate obstacles when producing clean historical renditions of type designs drawn from shabbily printed source materials.

John's observation suggests adding weight, or in this case 'ink' to historical type revivals. This is an inference I find particularly pernicious. It towers away from the craft of typemaking into the voodoo woods and whistles of mythical intellectualism.

It disappoints me that my work was not sufficiently apodeictic to be unquestioned by John. Apparently I had not made enough out of the fact that this particular cutting is unique in that supposition and second guessing takes a back drawer to the observation of hard artifact. Not missed was a rare opportunity 'not to wonder' where the ink ended and the paper began.

Fortunately the Lanston Monotype Machine Company had direct access to the original Caslon punches. Whatever adjustments Lanston made for the mechanics of the Monotype Typesetting Machine were duly noted in my writing in 'The Fount'. The ultimate test of this production in honour of the team of punchcutters in Philadelphia was 'The Caslon 337 Challenge' of which no expert including John, Mike Parker, David Berlow or Jim Rimmer were able to identify the original Caslon lines of type from the digital version.

Without my chart I have the same handicap.

I trust no one is suggesting Lanston should have mimicked English Monotype's failure with its production of Poliphilus. In their infinite wisdom, they not only honoured ink spread, they fossilized ink globules in their cutting.

NOTE: I have just been informed this is not how Poliphilus came to be. I will revise this when more is known.) For the moment allow me to quote Gerald Lange who writes me.
"But you mention Poliphilus. Later information on this is that it is not an exact replication with bumps and all but a "flavored" interpretation. Morison, once again, was lying. Beatrice Warde corrects this misconception in an article in the 1950s."

I must say to my taste this is just far too precious, and I am, as others, happy they discontinued that method of type revival. Certainly since then, and before, they have produced some beautiful work.

What we set out to accomplish here was to produce an exact replica in digital format of the original Lanston Caslon 337. Lanston did not clean the original for the hot metal version. It was successful. Lanston did not clean the hot metal version for the digital version, it is successful.

So the question is where you draw the line? The answer is you don't, the designer did. No ink, no globules, no repairs.

Maybe a little, aah . . . smoke!

Also I do not think John would seriously consider abandoning master source if he were to come across such a thing! Or seriously suggest improvements to 'the oldest living typeface'.

Perhaps he just does not like Caslon Oldstyle. That's alright. Some do, some don't. It has occurred when reading his lines about polymer plates that he prefers letterpress but is unable to formalize that distinction. I, for one, prefer wood over plastic.

In any event, Cobblestone Press produced under my direction very beautiful work using Caslon Oldstyle as its house face for many years. Robert Bringhurst has referred to me as one 'the experts' on Caslon therefore I speak with some authority. However I must clarify that my expertise is due to the frequency Caslon Oldstyle has found its way into my typographical designs and from the type channel of my caster. Some accounting should be taken of the time I have spent reviewing the original punch cutting production while working on the digital production and perhaps my little dissertation in 'The Fount' helped.

Alright, and I read a really thick book on Caslon once.

Anyway . . .

The erroneous conclusion of John Hudson about ink spread stems from his lack of hands on experience in letterpress printing and punch cutting. This experience is almost impossible to gain in this day and age. Although when John was in the plant he displayed focus. And for me, our time of exchange was far too short. I was busy, distracted and shortly afterwards Lanston relocated to Prince Edward Island. However John, and his partner Ross Mills were both attentive, listened and learned in spite of my idiosyncratic nature. I had earlier recognized in them the makings of great typographers. I prefer to take a measure of credit for inspiring them towards this industry. I recommend you visit their web site at Their work speaks quality.

But Ink spread is not inherent to printing on dampened hand made papers, it is merely the naked result of inferior pressmanship. Ink spread can just as easily be performed on machine made paper if that were desirable, which it is not. Ink spread falls into the same category as printed material out of registration, impressional show through, faulty back up, offsetting or thumb prints.

Ink spread is not an 'intended extension' of the punchcutters gravure.

I believe if a type designer intends to create various weights by using more or less ink they need their medication adjusted.

I believe Caslon would have cut his type bolder if he wanted it bolder.

And if the printer made the type bolder by his work it was his poverty, not his intention.

His type might be worn, poorly cleaned, perhaps even damaged. The paper might be difficult, the ink too long, the packing too soft, the machinery not adequate or his skills not up to potential. If a master printer requires bold type, the master printer sets bold type. That's why they make it.

Foundry type is set from a case to a stick, not knifed from a can.

To suggest otherwise is preposterous.

But as you will see it has been done before.

The task of the type designer is not to put ink on the type, it's to take ink off the type. Remember you read it here first!


UNLESS CIRCUMSTANCES are ideal large point sizes, bold type, or more simply put, large printing surface areas have special requirements. These often require a balancing of more impression, and or, more ink. The reason; Large surface areas tend to pluck, halo, seep through or produce the unforgiving heart break of offsett. These problems can, in some measure, be overcome with overlays, underlays, interlays, friskets, slip sheeting, offsetting sprays, gas flames, good fortune and sometimes a good ink Doctor.

Ink Doctors were hard to find in the industry. Now as we know, it's even hard to find the industry.

Solutions achieved by the above generally included some compromise. Hopefully no quality lacking enough to inspire customer complaint or criticism from ones peers.

In any event bold types are not often the object of historical type face revivals and a little ink spread is not proportionally as significant as a light text face.

That's the beauty of the system.

Letterpress with very little fuss can match sharpness in every detail to lithographic products. I made a series of reproduction proofs using Amatruda handmade paper while I worked with the very famous typographer, Errol Ettienne. Those proofs were used for corporate work including prestigious printings for MacMillan & Bloedel. Admittedly I was a fine pressman with a bag full of trade secrets. Normally I would not have picked Amatruda for pulling reproduction proofs, I would have used reproduction paper, a coated stock. However this particular project had severe time constraints. By using handmade paper for both allowed one operation to fulfil several purposes of which one of them was meeting a deadline. Incidentally, the typeface was Caslon 337.

When I delivered the proofs on Amatruda handmade paper I delivered Caslon 337, I did not deliver 'Caslon Crunch'.

Can you just imagine Caslon placing an advertisement in the 'Daily Cockney Smog'. Perhaps it would read something like this: 'Slovenly pressman required to perform inferior presswork in design department of the Caslon Letterfoundery, low compensation, high alcoholic perks. Eyeglasses: not only are they not required, they are forbidden. Job Description: Research technician for 'Multiple Master Ink' technology.'

Technically speaking all methods of mechanical printing demonstrate ink spread if viewed under a microscope. I presume John is speaking of excessive spread making it visible to the naked eye, 'weight associated with ink spread'. To imply these results enhance a type design, or that the designer intended his type to be bolder falls far from the mark. It would be like claiming that a wrong fount is not a fault but rather, a feature.

Perhaps I date myself using terminology like 'wrong fount'.

So, just why were smoked proofs the chosen medium for type designing founders?

And speaking of smoked proofs.

The Lanston patterns, in fact, were not smoked but lightly inked. However for any measurable ink spread on patterns of that size would be cause to pause.

Fortunately this is not important for this discussion. I just know how easily seemingly small oversights parrot their way into history.

Anyway back to the task at hand.

Many a hot metal type house thrived in the sixties hand setting reproduction proofs for advertising companies. We are talking crisp. We are not talking about printing with black berry jam on aunt Monica's pie crust.

I have an opportunity to tell you a story. Not about Caslon but Goudy Oldstyle. I had a repeat customer who had the same business cards, letterheads and envelopes that we re-printed year in, and year out. The same pressman at my shop handled the press work in all but the first printing.

Eventually the pressman went on to other studies. I was faced with performing the presswork myself although this was not my department. My usual crystal crisp clean presswork was delivered from the standing type form.

Much to my surprise the customer complained that I had used the wrong typeface. I had not re-set the type, I used the standing type form in the customer's assigned galley. This was without doubt the right typeface. The mark up on the docket verified the same.

This argument went back and forth until I reset the type, not in Goudy Oldstyle, but Goudy Bold. The customer reported they were now satisfied. That this was the type they had come to love. I remembered looking at the sign hanging above my press: 'The customer is always wrong'.

So was the previous pressman.

And so much for the ink spread theory.

Perhaps John prefers bad presswork or bolder types. One thing for sure is he would like aunt Monica's black berry jam and her pie crust. But surely John does not like 'Caslon Antique'.

That I do know about John. He is a man of refined taste and distinction and the 'weight' (pardon the pun) of his comments balance in favour of the accuracy of Caslon 337.

Now we get the 'Yes but . . .' crowd.

The yes but . . crowd says: 'Caslon looks best letterpressed on hand-made paper'.

So does Helvetica . . what's their point?

Self-respecting letterpress printers performing work for the 'design community' cringe every time they are asked to over impress type into paper. Stop it OK. Good designers should not order bad printing.

Alright, maybe I do have a few pet peeves. I've earned them.

Enough about ink spread already!

Optical scaling? Probably not!


Caslon Oldface
Caslon 337
Caslon Oldface

THE CASLON 337 CHALLENGE! 33,000 copies of the Caslon challenge were published but not one single type expert was able to dintinguish the difference. The end of my article explains how many lines of type were original hot metal, proofed, (Think ink!) and mixed along with lines of our digital font. The challenge is to pick those lines out.

So far not a single type aficionado has picked correctly a single line of the real Foundry Caslon from the digital version.

Now I don't drink but I have been told most people like their Scotch straight, that's how we serve Caslon 337.

Order a double Caslon and we will throw in a can of 'multiple master' black ink.

Remind me!


Postscript is a registered trademark of Adobe Systems Incorporated. Macintosh and TrueType are registered trademarks of Apple Computer, Inc. Caslon Oldstyle 337 and Lanston are trademarks of the Lanston Type Company / Gerald Giampa and his interests. See legals.
Caslon Letter FoundryCaslon OldfaceWilliam Caslon