Whenever I talk about rare books to someone outside of the world of book collecting, they're invariably familiar with the idea of a first edition. Most in the world of book collecting are familiar with the notion of first printings, many understand that books can have different states and issues. And many, even in the trade, get confused between issues and states. We thought we'd put together a short list of definitions of each. 

Edition

Technically and edition is all copies of a book printed from the same type setting. This is generally relaxed a little to allow minor changes to be included in the same edition. This is more obvious with non-fiction than with fiction as it's important to establish significant changes to the text as it becomes outdated. A work of fiction can have multiple editions, but the content will often by the same. It's important to note that a first edition can include copies of books printed over any time period (some books might be printed from the same type over a number of years).

Technically, therefore, should the plates for Darwin's Origin of Species still be available it would be possible to print more copies of the first edition.

Printing

Printing and impression are generally synonymous and refer to the copies of a book printed at the same time from the same plates. The first printing / impression would be those books that were printed in the first batch. The second impression would be those books printed from a second batch, which would usually be separated from the first batch by a period of time (weeks, months and sometimes years). Whilst different impressions might be identical to one another, and therefore indistinguishable, it is most common for printers to identify the various impressions in some way (number lines such as '1 3 5 7 9 8 6 4 2' or statements such as 'Second impression, 1955'. A single edition can, and often does, have multiple impressions.

It is important to note that any self-respecting bookseller will be referring to the first impression, if they omit the impression in their description; 'first edition' implies 'first impression of the first edition.'

Issue

Let's say that a book has been printed with a run of a 1000 copies. This is one edition with one impression. The publisher requests a few adverts to be bound in the back of the book. The book only sells 500 copies. Another order comes in a few months later and the publisher has a few different books to advertise and this time the new adverts are bound in. This becomes the second issue of the first impression of the first edition.

It's important to note that issues are groups of books within the same impression. offered to the market at different times; a second issue must have been offered after the publication date.

State

State is very similar to issue and often gets confused with it. The main difference is that various states are created during the production of the book and have no bearing on publication date. For instance, the printer is putting the book together and notices a typo on page 13. The printer corrects the mistake and carries on. The first 250 copies have the typo, the remaining 750 copies have it corrected. However, when the printer's about halfway through printing they notice that there's a whole paragraph missing. They correct that and print the remaining 500 copies. There are now three states in that impression. One has a typo and missing paragraph, one just has the typo, one has neither.

States and Issues

In the above example, there's a chronology between the states that sets the priority, but all three states hit the market at the same time so are within the same issue. If the 1000 copies were split between two different publication dates (with different prices for example), then you would have two issues, and each would have three states. At this point, the priority of the states becomes a little useless, so it's usually the issue that affects the value, as these are discrete chunks of an impression with a clearly defined chronology.

Variant

A book has various states produced before publication, and issues as distinct groups of books before and after publication. A variant, is where it's just unknown. It's a handy term used when it's not clear if the point is an issue point, or a state.

A Convoluted Story...Stop Reading Now if you Want to Retain the Will to Live

NOTE: Booksellers and collectors have been arguing for centuries on the point of issues and states. Even now, there'll be slight variations in definition. This is my take on it.

NOTE: This example is overly complex, but exists to provide enough room to illustrate the various complexities.

Let's say we have a book that's expected to sell 1000 copies. The printer starts printing. 200 copies in and a typo is noticed. This is corrected and the remaining 800 copies are printed. 

The book is bound and 100 copies are sent to the bookseller. 50 of the copies sent have the typo and 50 don't. The bookseller notices that one of the images is inverted and calls the publisher.

There are 900 copies remaining. Each copy has the inverted plate. The printer corrects these, but after doing 100 notices a missing page. The printer corrects the remaining 800. 100 more are sent to the bookseller. 20 of these have the missing page. Remember too, 50 of those 900 copies have the typo.

The publisher sees that the book is doing well and asks for an advert to be sewn into the book for another book they're publishing. They agree that this will be sewn into 500 copies, as the printer is a little annoyed and charging too much. The printer puts the advert into 450 copies of the book with the corrected page and plates, and into 50 copies of the book with the missing page. The bookseller buys the remaining 800 copies. The publisher breathes easy. The printer retires.

So, how many issues and states?

There are three issues as the book was issued to the bookseller in three groups, two of which were after publication date (they are all within the first impression). The second issue has the corrected plate, the third issue has the adverts (mostly).

The first issue was 100 copies and has two states. 

a) with the typo
b) without the typo.

All copies in this issue have the missing page, inverted plate and no ads.

The second issue was 100 copies and has four states, (arguably two states, each with two variants). 

a) with the typo, with the missing page,
b) without the typo, with the missing page,
b) with the typo, without the missing page,
d) without the typo, without the missing page.

All copies in this issue have the corrected plate and no ads.

The third issue was 800 copies and has eight states (arguably two state, each with four variants)..

a) with the typo, with the missing page, with the advert
b) with the typo, with the missing page, without the advert
c) with the typo, without the missing page, with the advert
d) with the typo, without the missing page, without the advert
e) without the typo, with the missing page, with the advert
f) without the typo, with the missing page, without the advert
g) without the typo, without the missing page, with the advert
h) without the typo, without the missing page, without the advert

A shed load of variations. Some of which might not even exist.

You pick up your book. It has adverts, typo, corrected plate and missing page. It's a third issue because the advert was only in the third issue, and is state a) 

You pick up your second copy. It has no adverts, corrected plate, no typo, and a missing page. It's got the corrected plate so must be second or third issue. It's either the second issue, state b) or third issue, state f). How can you tell which? Well, you can't. They are identical books. Even with all the printer's notes, you can't tell the difference.

Your third copy has no adverts, no typo, inverted plate, but no missing page. Unknown or mixed issue. According to the printer's notes, this can't exist as the inverted plate was an issue point and only present in the first issue. The missing plate was a state of the second and third issues. This is an anomaly. And either something changed by the bookseller, a mistake by the printer / binder.

How about looking from the bibliographer's perspective?

If this were examined retrospectively by a bibliographer, with no notes on how it was printed then the bibliographer might date the various issues based on the corrections and you'd end up with a different chronology. It might not even be possible to establish any issue points, and you'd end up with just different states. Let's take the obvious one, the inverted plate. If you have two copies, one inverted and one corrected, how could you tell which was corrected before or after publication? You couldn't. If you have a publication date, you could possibly date the book from the ads, which might feasibly split your impression into two issues. You might notice then that none of the second issue had the inverted plate, but still, you couldn't be sure that the inverted plate constituted an issue point, though it would be reasonable to assume that the inverted plate was an earlier state (as none of the second issue have the inverted plate). This would not imply an earlier issue though.

It's also logical to assume that typos are most often corrected rather than introduced. Copies with a typo are more likely to be earlier in the production.

Which is most desirable?

We know that the copy with the inverted plate is the first issue, we know that the typo is most likely the earlier of the two states. This doesn't really change the desirability of the book as it's part of the same issue, but some people might assign that chronology and prefer it.

If you have a corrected plate, typo, missing page and no ads, then looking retrospectively, you're going to still be assuming you have a first issue (remember, ads are our only way of delimiting the issues). You'll probably assume it's a later state as the plate's been corrected. But you'd be confused as the typo's still there and the page is still missing. So it must be early on, right? If we had the printer's notes, we'd know it was from the first 200 printed. But the reality is, that doesn't matter. To most collectors, the important thing is the books that were first published, i.e. offered to the public. This is why a lot of collectors ignore proofs and advance reader copies. Look at it the other way; imagine the book with the typo was the very first copy off the press, but it had the plate corrected and sat in the warehouse until the ads were bound in - still as desirable? Probably not.

Which is rarest?

Let's look at the printer's notes and try and put some numbers together.

Well, after the first issue, 50 copies with the typo remained at the printer's. They had the plate corrected, But 100 still had the missing page. It's possible that some of the 50 copies with typos were in that first 100 that had the plate corrected but not the missing page put in. Maybe only five copies.

100 further copies were sent to the bookseller. Maybe two of those copies with the typo and the missing page were in that one hundred. Technically, that would mean only two copies were a second issue, state a), and there'd only be three copies that were in the third issue, with the typo and the missing page. Put ads in one of those copies and the rarest is the third issue, state a).

Now, take it from the bibliographer's perspective with no publication notes. The bibliographer would see those 54 copies as one unit!

It's an extreme example, admittedly, but fun