2.2.1 The Late Upper Palaeolithic

So far, no sites, assemblages or stray finds from Perth and Kinross have been dated to the LUP, but considering that certain or likely material from this period has now been reported from most other parts of Scotland, it can only be a matter of time before finds from this period are made in the area. As shown in Table 1, the LUP consists of three sub-periods, namely the Hamburgian (12,700–12,000 BC), the Federmesser period (12,000–10,800 BC) and the Ahrensburgian (10,800–9,800 BC). Each are defined by a set of typo-technological attributes reflected in the material culture, most importantly as different forms of arrowheads (Ballin 2017). To date, the LUP in Scotland is recognised, and dated, solely on the basis of significant lithic types.

Two rows of seven sharpened lithics of various colours, from orange-brown to dark grey. The artefacts are photographed against a black background, with the largest lithics in the top row.
Detail view of lithics from Brownsbank and Howburn farms, including Upper Palaeolithic tanged points. ©️ HES

The Scottish Hamburgian is best known from Howburn Farm in South Lanarkshire (Ballin et al 2018), where several thousand artefacts (probably of Doggerland flint) were recovered, including many asymmetrical Late Hamburgian tanged points of Havelte type (a sub-grouping). On the west coast, a Federmesser period site, characterised by plain-backed points, was discovered at Kilmelfort Cave near Oban (Saville and Ballin 2009). Stray small Ahrensburgian tanged points have been recovered from the Loch Torridon and Inner Hebrides area (Ballin and Saville 2003), and a site from Islay may also date to this period (Mithen et al 2015). A single-edged point and several small-tanged points from Orkney indicate a LUP presence here (Ballin and Bjerck 2016). Two sites from eastern Scotland are also likely to be of a Palaeolithic date; these are characterised by unusually large blades (Long Blade industries; Barton 1998) which may suggest an Ahrensburgian date (Ballin 2019; forthcoming a). Sites along the Dee have yielded several probable LUP flint artefacts (Ballin and Wickham-Jones 2017). There is therefore no doubt that Scotland was settled, albeit thinly and possibly not continuously, over the later part of the Palaeolithic.

The key to understanding where LUP sites may be expected lies in the subsistence economy of the Palaeolithic groups and their high mobility. Obviously, groups on the Scottish west coast and that region’s islands must have exploited marine resources, and fishing is thus likely to have been part of the area’s economic strategy. However, the people of this period are generally defined as being highly mobile and primarily known to have exploited the migrating herds of reindeer which they followed and hunted across the steppes of the north European Plain, across Doggerland and up into what is now the Scottish mainland.

In order to consider the possibility of LUP sites in Perth and Kinross, it is useful to examine the known locations of LUP sites and findspots in southern and eastern Scotland. These sites are generally found in connection with rivers, such as the Tweed (Howburn Farm is located in the gap between the Tweed and the Clyde), the Dee and the Lunan. Rivers such as these are likely to have been important for many reasons: they facilitate travel and thus access into the Scottish hinterland, and are also likely to relate to the migration routes of animals such as reindeer funnelled along river valleys. Rivers and their valleys, such as the Tay, the Earn and the Isla, within the area of interest, would have provided varied resources and ecozones (for shelter etc).

With regard to site survival, identification of such early remains is problematic. In many places, LUP activity has left little archaeological footprint. There is an increasing recognition that the signature of exploratory and colonising activity expected in the LUP will be light. In addition, active geomorphological processes have served to obscure the record. In particular, the submergence of the Main Holocene Transgression has impacted on many of the lower fields along the present Tay estuary, which means that shoreline sites from the earlier Mesolithic and LUP are likely to have been disturbed and redeposited, as indicated by water-rolled lithic artefacts from lower levels along the estuary (Nicol and Ballin 2019; Dawson et al 2014). This suggests that, in the general Tay area including its tributaries, it may be more fruitful to search for sites dating to this period further inland. For example, on raised terraces along the waterways or around the shores of lochs (Howburn Farm type sites) which were places favoured by northern German and Danish LUP reindeer hunters (eg Rust 1937; 1958; Holm and Rieck 1992). Research into the latter category should consider the landscape changes as a result of agricultural drainage and other land improvement schemes from the eighteenth century onwards which have led to the disappearance of many ancient watercourses.