This slim, well illustrated, and reasonably priced volume seems quite likely to become a standard introduction to the Picts for the semi-popular market for some time to come, and Dr Cummins is to be congratulated on his achievement. The work is divided into eighteen brief but well focused chapters followed up by three appendices; two laying out the structure and calibration of the different versions of the so-called 'Pictish Chronicle' and one attempting to explain the catalogue of 'thirty Brudes' that appears in some versions of this king list.

Cummins, as he tells us in his first chapter, 'The Professor and the Pictophile', deliberately sets out to review the various problems of the Picts as they stand now, forty years after Wainwright's groundbreaking and influential book was published. The real strength of the present volume is that the author lays all his cards on the table. As each 'problem' is tackled the author sets out the primary evidence and clearly distinguishes it from the hypotheses developed to produce explanations, and whilst this reviewer finds himself frequently disagreeing with Cummins' preferred solutions to many of this problems he cannot but admire his evenhandedness and professionalism. Cummins' discussion of the problems inherent in using medieval documents as historical sources, especially in Chapter 3, 'Pictish Chronology', is amongst the best introductions to this thorny subject I have yet encountered. He attempts, successfully I believe, to make the reader aware of the nature of the relationships of different manuscripts to one another and to any hypothetical original, illustrating as he does so, how variant reading of the same narrative confuse interpretation and how the historian sets about dealing with these problems. For this virtue alone, the book is worth reading.

In Chapter 7, 'What were the Picts?', the question of Pictish origins and language are tackled. The author points out, quite rightly, that the Picts did not exist in the early Roman period when northern Britain was inhabited by a whole series of tribes such as the Caledones, Verturiones, Venicones etc. Pictish identity emerged as and when the peoples south of the Roman frontier zone became, effectively Romans, from a northern perspective. Cummins argues that it was the threat from the South which created the motive for a greater degree of unity amongst the northerners, and it could be further argued that the coining of the term Picti by Roman writers at this time served the need to distinguish between 'civilised' and 'uncivilised' Britons.

On the question of language Cummins, to my view correctly, gives no consideration to the non-Indo-European hypothesis. In a fairly lengthy examination of place names the author points out that the British elements, such as 'Aber' and 'Pit', are largely confined to the eastern lowlands and that most of the 'Pit' names contain a Gaelic element. He also notes that the distribution of the early Gaelic element sliabh, confined, outside of Wigtownshire, to a few examples in Argyll and in the upper straths of Tay and Spey, usually taken as evidence of an early settlement, looks, in fact more like residual distribution. He argues that the top end of valley systems and isolated peninsulas are where one would expect a language to linger on rather than originate. This argument, he further claims could apply to the distribution of Gaelic names south of the Forth, particularly in the upper reaches of the Clyde and nyeed systems. His conclusion is that the language of northern Britain was probably originally closer to an early form of Gaelic than anything else and that it subsequently picked up various British elements so that by the time of the Picts it was something of a mixed language. Certainly Celtic, but standing somewhere between Irish and British. To some extent this argument relica upon the only alternative hypothesis for Gaelicization being that the language switch did not begin until after Kenneth mac Alpin became king of the Picts and that Gaelic spread after 850 or not at all.

This reader is of the school that believes Pictish to have been a northern variant of the British language, albeit a conservative one and thus marginally closer to Gaelic than Old Welsh would have been. Nevertheless, Cummins makes some interesting points and it would be valuable to read the opinions of place-names scholars such as Simon Taylor and Richard Cox on such matters. Certainly the early spread of Gaelic is worth more consideration than scholars have allotted it. The possibility that the rise of a centralised Pictish state based on Fortrenn and Circinn, and producing the concentrations of Class II monuments and accounting for the late surviving Pictish place-names in this region, might coincide with a Gaelic conquest of some Highland regions on its fringes may be worth considering. Eighth century Atholl might, perhaps have possessed a Gaelic speaking population ruled by a Pictish sub-king imposed by Fortrenn.

Unsurprisingly Cummins is of the pro-Matriliny party. The persistence of belief in Pictish matriliny has become something of an Article of Faith amongst Pictophiles despite the absence of any historical evidence. Bede merely states that the female line was resorted to in time of doubt, common practice amongst the Welsh and the Romans, and of course the ideology underpinning the claims of the Maid of Norway to the Scots throne on the death of her grandfather. The Irish and Angle-Saxons preferred to follow the male line, however far back it had to be traced. Had traditional Gaelic custom prevailed in the thirteenth century, Alexander III would have been succeeded not by his grand daughter by Duncan III, Earl of Fife, indeed had Gaelic custom prevailed earlier in Scots history Malcolm II would have been succeeded by neithier Duncan I nor by Mac Beth but by Mac Duff. The cases of doubt mentioned by Bede were exactly such cases as that following the death of Malcolm II when British rather than Irish custom was followed. This is not the place, however, to engage in a debate that will probably run on for decades. Cummins is in very good company in his allegiance to matriliny.

In Chapter 17 Cummins presents us with a plausible, if partial, explanation of the symbols but his examination of Sueno's Stone is more fanciful. One of the frustrations, and, let us be honest, attractions of Dark Age History is that we shall never have all the pieces of thejigsaw puzzle, and perhaps, ultimately, this is where Dr Cummins falls down. Despite showing an awareness of the problems of the source materially, and explaining these problems very clearly, he is compelled, it seems, to find a solution to everything. Nevertheless, as an introduction to the Picts, in all their guises, this is a good book and the author always allows us to see when his own ideas deviate from the consensus. For serious, if at times over cautious scholarship, one should turn to the essays in the Society's own Pictish Panorama, but for a good read or an introductory gift to a fledgling Pictophile Cummins' volume would be an admirable choice.

Alex Woolf