This book sums up, for both the general reader and for expert linguists, everything (for all practical purposes) that is currently known about Irish language Ogam. Chapters One, Two and Three introduce and explain how Ogam originated and how it was used. This is general information that everyone interested in Ogam in whatever language should know. Chapters Four, Five and six, on the other hand, are for persons with some expertise in linguistics. These chapters build on the comprehensive collection of diagrams and photographs of Ogam inscriptions (as updated by recent finds and certain more critical readings) that are contained in the first volume of R A S Macalister's two volume work Corpus Inscriptionum Insularum Celticarum (1942, 1949), which is the standard work in the field. After this, Chapters Seven and Eight discuss, for both the general reader and the expert, what is known as Scholastic Ogam. The book is rounded out by two appendices, extensive notes and an exhaustive bibliography.

In discussing the origin and use of Ogam, McManus refreshingly presents simple and down to earth answers to questions which are notorious for arguments that generate more heat than light, and for theories which fiequently incorporate unbridled flights ofimagination. "[Ogam] is not", he writes, "the creation of a dilettante whiling away his leisure time toying with ciphers, but a carefully planned and co-ordinated writing system designed as a vehicle for a language with a phonemic structure of its own". To demonstrate this, McManus divides consideration of the origin of the Ogam alphabet into two areas: its signary and its internal structure.

"The characters", he explains, "of the Ogam signary are not alphabetical graphemes; they are integral parts of a linear code which by its very nature is inflexible, and is clearly unconnected in origin with alphabetical writing". The origin of Ogam therefore has to be a tally system. This is a system in which successive units are counted by making marks on stone or wood. However, tally systems for effectively dealing with any more than two or three marks need to incorporate three principles: ordering, grouping and abstraction. Thus, while a long row of tally marks might be in order, it would be quite unwieldy to handle. To facilitate counting, such a row would need to be divided in some manner into groups. The next step would be to use some other symbol, in lieu of a number of tally marks, to represent such a group. Roman numerals are a good example of a well developed tally system. Here, five single tally marks are abstracted into the character "V". As a result, the sequence "VII", for the number seven, consists ofthe "V" as the rank marker (for five units) plus two marks. The designers of Ogam, instead of devising symbols for a rank marker, ingeniously decided to employ a system of having up to five marks above the stem line, five marks below the stem line, five marks diagonally across the stem line and (originally) five notches on the stem line itself. While this method limited the number of units that could be accomodated, the twenty position sequence thus obtained was sufficient for their needs. That such a sequence might also be used for telegraphic or cryptographic purposes did not diminish its utility. Moreover, by locating the inscription along the edge of a standing stone, an Ogam inscription was able to use the edge itself (the arris) as the stem line.

In dealing with the internal structure of the Ogam alphabet, McManus explains that the distinguishing characteristics are: "the inventory of phonemes to which it caters, the sequence in which they are arranged alphabetically and the names which they bear". Failure to realise this intention of Ogam's internal structure is what has led to the innumerable frustrating and convoluted efforts to align the Ogam alphabet with the Greek and Latin alphabets. The designers of Ogam understood, as did Greek and Latin grammarians, that letters rather than sounds were the smallest parts of speach. They also realised that there was a difference between vowels (or vocales) and consonants. Moreover, as the Latin grammarian Varro (whose era was contemporary with the development of Ogam) had pointed out, the identification of consonants which are semi-vocales began with "e", such as "ef', "el", "em", "en", etc, while those which are mutae ended with "e" - "be", "ce", "de", "ge", etc. The terms used to designate the letters were therefore of the utmost importance. Then, as now, the memorisation of the letters of the alphabet was propaedeutic to learning to read and write. The terms originally adopted to designate Ogam letters are, as McManus points out, words which indicated sounds of Primitive Irish as those sounds were perceived by the designers of Ogam. Because of this, these letter names do not necessarily have the same phonetic values that are attached to them in the alphabetical lists of Scholastic Ogam in Mediaeval manuscripts. Scholastic Ogam values are the result of a revision of Ogam in the Old Irish period to make it compatible with the Latin alphabet. It should be noted that the original Ogam letter names were not all the names of trees or plants, as has sometimes been suggested.

The linguistic upheavals which accompanied the transition from Primitive Irish to Archaic Irish to Early Old Irish to Old Irish (of the seventh century), when the original use of Ogam went into decline, resulted in four of the original letters becoming redundant. These letters, which were re-aligned in Scholastic Ogam with phonetic values of H, Q, NG and Z, represented sounds that did not belong to Primitive Irish and, in the case of NG, for which there was not even a comparable letter in the Latin alphabet. In Chapter Three McManus traces both the changes of phonetic value and the modification in names which accompanied the redundancy and re-designation. In Chapter Seven he again picks up the story with the addition of the five new symbols: EA OI, UI, IO and AE, collectively known as the forfeda. These he describes as "having no internal consistency and little or no relevance to Irish". However, each forfid was "now pressed into service to denote a dipthong or a digraph beginning with its initial vowel and was thus distinguished from the letter name in [original Ogam] . . which denoted the pure vowel".

Having demonstrated that the original Ogam alphabet was designed to record the sounds of Primitive Irish (for which there was no other written script), it logically follows that its use was confined to situations where something had to be written out. In essence its use was similar to that of the Romans in carving inscriptions in capital letters on monuments. Hence it is a convenient distinction to refer to the original Ogam alphabet as monumental Ogam, and to the later Scholastic Ogam as manuscript Ogam. In examples cited from the Early Irish law tracts, it is indicated that Ogam inscriptions were persuasive evidence in legal proceedings for in heritance and ownership of land. Eventually the word ogam became a generic term for writing. Then, as McManus says: "At some time in the seventh century Ogam fell into decline in its capacity as a monument script and, from the eighth century on, was replaced by conventional script in the form of Irish semi-uncial". However, this demise of original Ogam did not mean that inscriptions in semi-uncials continued the Ogam tradition in a superficially modified guise. There was an entirely new beginning: "not only in script and orthographical conventions but also in distribution and the general choice of recumbent slabs as opposed to the more common standing pillar of the earlier period".

Chapters Four, Five and Six contain, as previously mentioned, quite detailed information on the names used in the inscriptions, how it is possible to date inscriptions by analysing their grammar, and the difference between Irish inscriptions and Ogam inscriptions in Wales and on the Isle of Mann. Regretably, McManus does not discuss Pictish Ogam inscriptions. This is quite understandable in view of the type of analysis being undertaken, since there is no equivalent Pictish language basis from which to work. However, since Pictish Ogam appears to have been borrowed from Irish Ogam, McManus' explanation of exactly what was being borrowed is a very necessary first step towards the eventual translation of Pictish Ogam.

Bill Grant.