Following are notes for a talk that I had planned to give in Aug, 2004:

     As a dealer and bookseller in Asian-African series I often critique references in order to advise customers what to buy, and sometimes get requests from authors for advice. Numismatic cataloguing is an age-old and international practice. Methods and goals differ among mass market, academic, trade, and self-published works. But standardization is the inevitable way of things, and discussion among users, authors, and publishers over the details can help preserve the best of the past amidst the opportunities offered by changing technology. What follows is a sort of advice-to-authors compilation. I have tried to identify options, and state my own preferences in balancing user convenience, ease and cost of production, and appropriateness to topic. Hopefully you will be able to identify topics I have missed, yet more options for those discussed, and defend choices which I have shortchanged.
     My comments are directed mainly at authors of specialized catalogs, rather than works in economic or numismatic history, or omnibus catalogs such as the SCWC, though the latter will be invoked as example due to familiarity and to the spotlight cast by Krause's recent global change in certain format elements.
     Three general observations: 1) Authors generally write with other specialists in mind, not realizing that generalists or complete outsiders account for 90% of the sales of their works. Simpler is better. 2) A good numbering system is important to the acceptance of a new catalog. It is not an afterthought; it can make or break a work. Simpler is better. 3) Seek detailed advice from collectors, dealers, and publishers during the production of your work.
     Users of numismatic references generally want these elements: Historical (brief) and numismatic (more) background information; a catalog that is comprehensive by type and more detailed by variety than earlier works, well-illustrated, with a numbering system and valuations. A finding list if the material requires it.

Elements of a Reference work (Terms I've used)

Specs: Characteristics and differentiators of an item, which may include: Catalog #, Type/subtype/variety, description of key elements, government or reign, issuer, mint, date or range, denomination, size, weight, composition, mintage, edge type, die axes, designer / engraver, notes on variation range where such variants are not catalogued, translation of legends, brief description of theme, brief description of legends, pictorial elements and dividers and their relationships, cross-references to other works. Some specs may be more effectively presented in an appendix. Some can be more concisely conveyed by using symbols rather than words.
Illustrations: may include obv photo, rev photo, rubbing, drawing, enlargements, drawing of detail area, side-by-side comparisons; legends copied out for clarity or transliterated; graphic highlighting elements such as circles & lines.
Valuations: Market prices, issue prices, rarity ratings.
Background: All elements of numismatic & economic history, including discussions of metrology and denominations, mints or issuers; collector-oriented tips, anecdotes; comments on rarity or availability.


     Why you chose to write on this subject, enjoy it, or came to collect it. Discussion of previous literature & how this book compares. Points of difficulty in the field & why you have chosen certain methods of analysis or presentation. Comprehensiveness: Have you tried to list everything known, the holdings of a particular collection, a selected sampling? To what depth of variation? All are valid choices, but the reader wants to know your criteria. This is the place to frame your series historically, mention how it relates to better-known series, or how a study of numismatics in this area aids understanding of history, religion, art, or other aspects of culture. Advice, encouragement & warnings to collectors new to the series. If you present new theories or a substantial number of previously unpublished listings, mention here. A decision to buy, or use, a book is often made after a flipping through of the text, and a reading of the introduction and table of contents.

Catalog Key or "Abbreviations & Definitions"

      If you use many abbreviations or specialized jargon such that a novice user would have difficulty with the catalog, consider boosting visibility of this section through placement or graphics, or repeat it before each division of the catalog. Consider using a table format, and translate these terms into several languages - easier than making the whole work multilingual, yet broadens your audience. A discussion of how you differentiate types from varieties may be helpful if this is not obvious from the material or handled in the introduction.

Catalog structure

     How a catalog is ordered - alphabetically, historically, geographically, a combination thereof, or according to traditions established by earlier authors in the field - is the first major decision an author faces. My own preference is for a method that best illustrates how the items came to be, or how they relate to their culture, even if this is at the expense of findability within the catalog.
     There is a serious difference between a phenomenological or "pigeon hole" (PH) format, and a historical or geographic (HG) format. The first lists items by their appearance or particular specs, the latter according to known but less visible factors such as place or sequence of issue. A PH format may be the best achievable for Westerners cataloguing Oriental series, or token series where issuing entities are largely unknown. When the body of knowledge advances to the point that a HG format is possible, the PH structure should be preserved as a finding-list within the work. An example would be Oriental charms which are grouped by visual characteristics such as shape or pictorial elements in Western works, but by intended use in Oriental works. Numismatic items with Chinese characters are now generally indexed (and catalogued as well when information is lacking) by stroke count, using Pinyin rather than Wade-Giles.
     An arbitrary system is better than none at all. In cataloguing Chinese open-worked charms, most of which display animals or plants, I listed roughly in phylogenic order. When neither historical, geographical, alphabetical or other considerations argue for a particular sequencing, my preference is to move from commonest to rarest.
     Don't divide your subject unnecessarily. A smooth progression along historical, geographical, or alphabetical lines is best, keeping in mind that most users will never understand the series to a depth which makes divisions along esoteric lines sensible. (Jen's CHINESE CASH, arbitrary division into main and variant sections; Craig Coins of World 1st ed. listed Asia-Africa separately).

Layout & Display formats

     A good layout makes it easy for the user to pluck information from a reference and move on, which is what the vast majority of users want to do. My own preference is for works which make full use of font variations such as separate fonts, bolding, and sizing, and positional elements such as indentation and centering - to whatever extent the complexity of the subject mandates. Columns or even a gridded table will ease the eye's burden for listings which include a large number of specs. It is also important to repeat category headings with each double page spread.
     As collector and institutional websites themselves evolve toward catalogs and authors make choices based on translation from print to web or vice versa, bear in mind that desktop publishing programs are much more versatile than the crude and still evolving HTML language in creating eye-friendly layouts. An online catalog's advantages include low cost and ease of update, while a printed catalog's include more universal access, ease of use, ease of comparison (object to image), and permanence. Not considered here, but a useful topic, would be the ease of translation to online format of various print formats, and conversely, ways of designing online catalogs to facilitate compact printouts with satisfactory graphics.
     The Krause-Mishler SCWC series is a special case, as it attempts to organize a tremendous amount of varied material within a consistent framework, and to do it with a minimum of human input. KM has moved from a layout-rich "old" format to a more atomic "new" format, which allows discreet listings to float freely and determine their own column and page breaks. This is a great effort saver for the publisher, but creates a less user friendly product. Consider these layout elements which have been lost: catalog #'s in columns, denomination and mint categories in bold type, commemorative theme centered below illustration, spec categories on separate lines. In non-Western series where the mint name most readily distinguishes similar issues, this loss of layout elements can seriously impair usefulness.
     What specs you include will depend on your subject matter. Generally it makes a clearer presentation when all listings of a given category share specs in common, to give these specs below their category heading rather than repeat them with each listing.
     Why describe visual elements of an item if you have a picture on the sme page? Token references often repeat legends in the catalog which are easily visible in the illustration. A logical breakdown of type and variety can be conveyed with layout elements such as bolding, sizing, and indentation, so that the item description need only reference characteristics which differentiate that item from those coming before and after, especially elements not readily visible in the image. Historical and anecdotal information particular to that one item can also be given in the item description.
     When your subject material uses scripts such as Indian and Arabic, which will be unfamiliar to many catalog users, and especially when strike or wear often render portions of the script hard to read, consider including in your catalog a transcription (copying out in the style used on the coinage), and possibly transliteration (rendering it in Western letters), and translation. For Chinese, a transliteration is essential, while a translation is often meaningless.
     Illustrations within catalog is almost always preferable to a catalog / plates division. The main virtue of splitting is a savings in printing cost due to greater compactness, and use of coated stock (for good photo reproduction) in the plates. For smaller works, the savings may be marginal, and well offset by user convenience. When a series consists of many similar items, separating plates and using a large sized page format allows the user to make comparisons more easily. An example is the Burger work on Qing Cash where foldout plates several feet in length display coins that may differ only in the angle of a single stroke. Even here it is helpful to illustrate type or class specimens in the catalog portion, and consider the plates a sort of appendix for the subset of users who pursue the subject to this level.
     One technique seen in Japanese catalogs and seldom elsewhere is to use circles, lines, and arrows to highlight a coin detail and link it to an enlargement or defined icon next to the actual illustration. The purpose is to draw attention to a key difference between similar varieties. Western works generally use verbal descriptions which consume more space, require interpretation, and are language-specific.
     Consider Michael Mitchiner's Oriental Coins and their Values series for layout. Each chapter or division begins with a political and numismatic history with dynastic lists, followed by a catalog. The catalog format is three part: Catalog # in one column, specs second column, and photos third column. Bolding and italics are used to differentiate categories of specs. Dynasties under a given heading are numbered. A larger font size for category headings (usually ruler's name) would help readability and a summary of hierarchical headings (Dynasty / Ruler) at the top of even numbered pages would save paging back to orient oneself. But generally it is a user-friendly, consistent format under which a great deal of different material is organized. The downside is cost: coated paper used throughout, and looseness in the layout of either text or photos to maintain a correspondence.

Numbering systems

     The only modern catalogs which routinely omit numbering systems are Chinese works on cast cash, probably because saying or writing four single syllable characters is nearly as easy as and more informative than reciting a four digit number, and the authors care only about Chinese-speaking readers.
     A numbering system is a necessity, yet too many authors do it badly. Ideally a catalog number should be compact, easy to remember, present types and divisions thereof in a hierarchical manner, allow logical interpolation of later discoveries, and sort properly by computer. Yet no such ideal system exists; all are compromises. Tradeoffs must be made based on your listing format, comprehensiveness, and the likelihood of future issues or discoveries. An author's best strategy will depend on whether he is writing a pioneering work in a field where new additions are expected, or a work where mainly variants, rather than types, will show up in time. My own preferences as a dealer who uses standard catalogs (in each field) to organize stock is for a simple 1+ pattern with the first type presented given #1, and so on to the last item in the book. This allows the reader to quickly measure the scope of the field, to organize his collection by the book, and quickly determine what is missing. It also sorts well by computer.
     Opposed to this is what I call "suitcase numbering" where the number itself carries information about the type rather than simply representing it. Suitcase numbering is found quite often in token references, where for example 147CVA-5 represents the second issue of the 1st Cavalry Division (Airmobile), value 5 Cents. This is a good "working" system for an author organizing a huge field where new material is continually being discovered, but when final draft time comes, why not just number consecutively leaving gaps between issuers for later interpolations? A lesser form of suitcasing in coin works is starting the numbering over with each chapter division, which necessitates (usually) alphabetical prefixes to the base number to distinguish the various 1+ runs within the book. Such a system is difficult to sort, manually or by computer, can lead to miscommunication, and works against the vanity factor of having the author's initial associated with the number system. It is a firm tradition within the field to use 1+ numbering for each political entity, thus when a single work catalogues multiple entities, a conflict arises. Within the background section of the work, using alphabetical prefixes to distinguish the run of each entity (as the issues of neighboring or interacting states may often be discussed together) may be helpful, but printing such abbreviations as a part of the catalog itself is unnecessary.
     A second consideration in a numbering system is hierarchical ordering of Type, subtype, variety. My own preference is for alphabetic suffixes over decimal suffixes. For the tenth variety of type 2, which is more compact and sorts properly by computer, 2j, or 2.10? The whole number (2) should be reserved for the general type while 2a and 2b describe to varieties. This allows referencing a coin as a general type without specifying details. If decimal suffixes are used, 2.1 and 2.2 should be the first varieties, as using 2 for the type and 2.0 for the first variety is confusing.
     What if you have more than one level of variation? You may wish to broaden your definitions of "type" entirely to avoid this level of complexity. Or you may expand the 1+ numbering to run sequentially through 1a, 1b, 1c, 2a, 2b, etc., using a header to group "Type 1" coins together with a descriptor of its characteristics. However, using a second level indicator can help illustrate the hierarchical pattern, and make your analysis of the material clearer. Consider the combined use of alphabetic suffixes, decimal suffixes, and dash suffixes (which sort between whole numbers and decimals).
     The most difficult challenge for a numbering system is interpolations. In a field where many new finds are to be expected after publication, my preference is to leave generous gaps in the 1+ sequence, between each issuer, series, etc. In a more studied field, I would use a gapless 1+ sequence and number interpolations with alphabetic prefixes, A2 being a type later found belonging between 1 and 2 (and hope nothing between 1 and A1 shows up!). Better yet, save those decimals for such a purpose and call it 1.5. A good numbering system will allow different users who have studied your system to independently assign the same new number to an unlisted type.
     If your work is primarily background and includes a catalog mainly to illustrate certain points, why invent a new numbering system? Refer to the items by KM numbers or the reference most common in the field. I'm not aware of any publisher who discourages such use.
     If your book is a more comprehensive treatment in a field with a pre-existing standard work, or a new edition with a revised numbering system, it is best to give the older reference numbers within the body of your catalog, if not directly adjacent to the new numbers. To the degree that your own catalog is more comprehensive, detailed, recent, or easily available than the older reference, a better case exists for a "clean sweep," relegating the old numbers to a separate concordance in the back.


     Valuations are very important to the salability of a work to the collector market, but apparently of negative prestige value. Museum catalogs and academic works typically omit them, while collector-oriented works almost always include them. A common compromise is to list valuations in an appendix, or in a separate insert. The latter solution also allows revisions of the insert alone, as values inevitably change over time. Often authors who are quite well qualified to catalog a series have little idea of market values. A solution is to farm out the job to an active collector, or dealer, but better to get multiple opinions. Be sure your values contributors are not trying to manipulate the market in their own favor. Some discussion of typical grades found is useful, and valuation lists should indicate what grade is being valued unless they specify different values for different grades. Listing overly precise values or values for several grades when the market has not been carefully studied is a disservice to users, as is giving only one price for "average condition" without specifying what condition is average.
     Rarity ratings may be an adequate alternative to actual market variations, but will not satisfy some prospective buyers. Giving a formula for converting rarity ratings to valuations is misleading, as rarity is only one factor in market demand.

Appendices & Tables

     These can be used for material that is basically specs or background but either more conveniently presented or used apart from the catalog, or of lesser or specialized interest. Glossaries (Language, script, or pronunciation guides), historical timelines, concordances with other references, dynastic lists, lists of mints or locations, discussion (or catalog) of forgeries and imitations, ownership of specimens illustrated.
     In catalogs of particular collections, weights are often given in an appendix rather than with listings to avoid implying that specimen weights are average or typical for their type.
     An appendix can be used to cover numismatic or semi-numismatic items that are strictly outside the scope of your topic, but may interest collectors. Advanced collectors are always looking for something familiar but new. A work on official coins and banknotes might include tokens, medals, chits, charms, etc. as an appendix to give readers a taste of these areas without having to maintain standards of comprehensiveness or rigor in description.

Other elements

Table of Contents: If your book is not in English, include English translations for the chapters listed here. It will help sales to American buyers.

Index: It is very common for numismatic works to lack an index, or include a poor one. I strongly urge authors to spend more time with their indexes, keeping in mind that the great majority of users want to quickly find a fact, a spec, or a valuation and then move on. Farm out the job if necessary.

Map: A good addition to even a simple work where it helps illustrate or explain the relationships among mints, issuers, and the like. Book buyers like maps.

Bibliography: My own preference is for less rather than more. Authors tend to be over-inclusive to demonstrate the breadth of their research. Stick to major sources, and those dealing with controversial points. Abbreviate the names of works frequently referenced in the text.

Language: Even if making your reference fully bilingual or multilingual is impractical, translating the introduction and key to catalog listings into a language understood by local collectors of your series will broaden its market and help bring together homeland and Western collectors. If your book is not in English, translating at least the Contents will help sales to American collectors.


     There is rarely any profit in publishing a specialized numismatic work. Finding a publisher, or a printer, are essay-length topics themselves. I could mention one or two firms to avoid because they do not work well with dealers or get good distribution. I am impressed by a print-on-demand service with outlets in Europe and North America, such as Trafford Publishing.
     People DO judge a book by its cover; don't economize here. Use a bold, color graphic and preferably a color background, even if it means color photocopying on card stock.
     Steel wire and plastic loop bindings which exceed the width of the book itself, along with looseleafs in binders, are more difficult to shelve and to identify once shelved. Many booksellers and particularly distributors will avoid books with such bindings, regardless of merit. Positive aspects are low cost for small or sporadic print runs, and the ability to lie flat without spine cracking. Squareback, glue-binds, as done by many auction firms, are inexpensive now for relatively small print runs. At some added cost, pages can be sewn in signatures first, providing a longer life of heavy use. Singapore is often mentioned as a cheap print location. ISBN numbers, which were seldom used in numismatic books in the past, are now critical for getting your work listed and sold by, Barnes & Noble, and many other large retailers.
     The size of your book affects its mailability internationally. The maximum size that will fit in a flat-rate envelope is 8.25 x 10.75 x 1.25 inches. These special envelopes provided by national postal services travel at a fixed rate not dependent upon weight, and make it more practical for sellers in one country to mail to buyers in another.