Sources & methods (2): Directories (1)

Database size. Cleaning. Categorisation.

Introduction

Sources & methods (2) discusses collection and analytical processes applied to records listed in commercial directories during Bournemouth’s Victorian period.

Sources & methods (2): background

Commercial directories have survived in a haphazard manner, scattered across libraries and online portals. Several publishers competed in this market during the Victorian period, Kelly emerging as a leading participant towards its end. Little has emerged as to the methodologies applied by the publishers. They perhaps relied on local agents who worked in a context that enabled them to know well their communities, for example, clergymen. Publishers may have copied from each other, some information, therefore, not referring to the same time. Mistakes sometimes occurred as to names and their spellings. Nevertheless, the collection assembled for this study tracks Bournemouth’s trading community from 1849 to 1898. The volumes found provide the situation more or less for every five years over this period.

The database contains almost 20,000 records (including 1905 and 1911). About 12,000 records concern the Victorian period. The number of records contained in each sequential volume always increases, since Bournemouth’s population and economy grew at a steady rate through the Victorian period. For example, in 1849 a directory contained 45 names comprising Bournemouth’s economic sector. By 1871 this had increased to over 300 names. At the end of the period, the number had increased tenfold to over 3,000. For the most part, each record refers to a business, but, towards the end of the period, the directories also included Borough inspectors of varying nature. In time, company names began to appear, but, the majority of records refer to individuals.

Sources & methods (2): data cleaning

Data collection for the directories may not have followed such a rigorous and centralised format as applied to the census. The balance between details supplied by the business owners and third party agents remains unclear. Coverage levels will always remain unknown. Nevertheless, since the numbers increased over the period, when Bournemouth grew at a regular rate, they must have reflected reality to some extent. Data cleaning leads to standardisation amongst variations occurring for business descriptions. An obvious case consists of ‘coal & corn merchants’ and ‘corn & coal merchants’. The difference seems more attributable to random variation rather than a reflection of emphasis within a business. Thus, the descriptor for all such businesses becomes ‘coal & corn merchant’. Coal merchants who appeared not to supply corn remain so described. In a similar fashion, the ‘butchers’ include so-named businesses but also ‘pork butchers’, ‘colonial butchers’, ‘beef & ham shops’ and ‘meat purveyors’.

Sources & methods (2): categorisation

Analyses require different levels of granulation. Categorisation of the records provide these. At the lowest level, the cleaned records resulted in over 400 identified commercial activities. This enables analysis to address tailors or painters or solicitors and so on. This supports drill-down and trend analyses for the activities, but some questions require a wider focus. Two levels of ‘clumping’ have evolved for this purpose. The first packs the businesses into one level of meaningful categorisation. This produces 45 categories. For some perspectives, this remains too granular, so further grouping has resulted in ‘super-categories’. This reduces the number of groups to fourteen. Thus, this programme of data management results in support for analyses at three different levels. At any time, the analysis can move between the levels, but also down to the original record wording.

Sources & methods (2): example

As an example of this process, take George Joy. In 1871, he had a game dealership, located in the Arcade. By 1875, he had moved, still selling game, but now having pork as his main line. Thus, he fell into two primary categories: butcher, game dealer. At the second level, however, both of these fall into the categorisation ‘meat & fish’. In addition to these two activities, this group includes fishmongers, poulterers, ham and beef shops as well as sundry other related businesses, for example, egg dealers. At the final level of categorisation, the fourteen ‘super-categories’, George Joy’s business falls into the group labelled ‘food & drink’. In addition to meat & fish, this group includes: bakers, confectioners, dairy, fruit & vegetables, grocers & provision merchants, wine-spirits-beer shops. In 1898, almost five hundred businesses fell into this group, led by grocers & provision merchants, then meat & fish dealers.

Sources & methods (2): additional points

The directories contain no other information on businesses than their name, that of their owner or manager, and some details about address or location. No quantitative information exists in this source. This precludes analyses based on selling space or anything related to turnover and profitability. At the fullest extent, the commercial entities stored in this database refer to the overall area ‘Greater Westover’. An analytical construct, it makes allowance for how Bournemouth’s footprint grew during the Victorian period. The database contains enough locational detail, however, to support analyses by different parts of the town as well as the suburbs which flourished in the period. It also contains records relating to the old rural settlements which came to form part of Bournemouth’s economic and social hinterland.

References

For comments, please get in touch here. Good entry into directories here.