2. The connectedness of internal and international migration

International migration is usually seen as a distinct and separate process from internal migration, but, although recent influential work on migration and mobility has embraced most aspects of human movement,Footnote ³ only rarely have the two been explicitly linked together. Moch's seminal work on migration in western Europe since 1650 deliberately considers movement both within and between European nations,Footnote ⁴ and Hoerder's massive work on global migration explicitly states that his first goal is to ‘describe and analyse migration from the local level to the continent-wide and global’.Footnote ⁵ He begins by recognising the ways in which both global and local movements may be encapsulated within a single life course: ‘It [migration] begins with departure out of parental households and ranges as far as transcontinental or trans-oceanic moves – a geographic scope that might be one and the same move of a man or a women’.Footnote ⁶ However, while recognising these crucial links, the connections between different types of movement are rarely considered explicitly. Relatively recent recognition of the need to link the global and the local in historical migration studies comes from Feldman, who argues that ’we might extend our understanding of both internal and international migration if we examine both within the same conceptual field’.Footnote ⁷ He suggests that too often local migration has been explained by a range of social, economic, familial and cultural factors, and has been assumed to be relatively benign; in contrast international migration has been examined mainly within an economic framework which assumes that such long-distance moves were undertaken almost exclusively for betterment and that they mostly had much more serious consequences. Following Moch's argument,Footnote ⁸ Feldman argues that such a distinction is not sustainable. One of the few studies to examine explicitly the links between internal migration within Britain and emigration from Britain was that of Baines.Footnote ⁹ This county-level study used aggregate data to estimate the number of migrants who emigrated directly from urban areas, and the proportion that had first moved from rural to urban areas. He concluded that most emigrants from England and Wales in the period 1861–1900 came from urban areas, with only about one fifth having moved from rural to urban areas prior to migration. Some of the same themes were addressed by Knowles.Footnote ¹⁰ She examined the migration histories of emigrants from Wales to the USA in the nineteenth century by linking their migration paths within Wales prior to movement with their overseas movement, and their subsequent migration and engagement with the US economy at their destinations. Knowles places particular emphasis on the role of both religion and occupational skills (in the charcoal iron industry) in constructing these migration pathways. Such complexities are even more obvious if one considers migration in other parts of the world. Thus, in the context of African migration in both the past and the present, Adepoju argues that ‘conceptually, international and internal migration are complementary and can indeed supplement each other’,Footnote ¹¹ and in many parts of the world (including continental Europe) a short local move can easily cross a national border.

3. Constructing a longitudinal migration database

The potential significance of linking internal and international migration is further enhanced when migration is viewed over the whole life course, rather than as a series of discrete events, and this can be illustrated through selected examples drawn from the migration database collected some 25 years ago by Jean Turnbull and myself.Footnote ¹² The aim of this research project was to collect a large number of (more-or-less) complete migration life histories of individuals born between 1750 and 1930, and who had lived in Britain for at least part of their life. This was done by contacting 80 local and national family history and genealogy societies widely distributed across Britain, with a request to publish in their various magazines details of our research, and with a request for volunteers to participate. This would entail completing a detailed migration life history form for every ancestor for whom they had relevant information. All bar 17 randomly distributed societies agreed to help with the research, 2,420 individuals agreed to participate, and a total of 28,698 forms were distributed. The form was quite demanding as it required respondents to provide details of every move they could identify over the whole life course of an ancestor whom they had researched, including addresses, marital status and family structure, occupations with dates for all vital events and moves. They were also asked to provide (where possible) a reason for a move and to note the sources used so that we could check that data had been generated from a viable source rather than speculation. In total 17,161 forms were returned and, after careful checking for completeness and consistency, 16,091 forms were deemed usable. These provided information on 73,864 different residential moves and these data were coded (where appropriate) and entered into three linked databases: one organised by individuals, one by moves and one by occupations. This provided a unique and flexible set of data that could be analysed in a variety of ways, and which to some extent mirrored the longitudinal data available in continuous population registers.Footnote ¹³

All historical data has its limitations and, as with most evidence, the databases have some inevitable biases and omissions that must be taken into account when interpreting the data. Three key issues are highlighted here: validity, representativeness and demographic biases. It may be suggested that ‘amateur’ family historians would be less skilled at identifying and interpreting historical sources than professional academic historians. However, I have no evidence that this is the case. Family historians invest an enormous amount of effort in tracing their ancestors and have a personal investment in the project. In contrast, academic researchers usually have no intrinsic interest in individuals but are usually seeking to amass a large data base to identify broad trends. They may thus invest less time in following up individuals in a range of different sources. Moreover, many family historians have themselves gained a high level of skill through diploma and masters courses, and research for this project was carried out prior to the ready availability of family history data on-line, and thus family historians had to come to grips with a wide range of original sources in a variety of different locations.Footnote ¹⁴ In collecting data we required respondents to provide details of the sources they used so we could check the validity of the material, and the quantity of detail required by the forms would have deterred many less-skilled potential respondents. Although all historical material may be subject to error, we were confident that the data provided had a high degree of accuracy. Undoubtedly, the most problematic information to collect was about the reasons for a move which, in many cases, can only be inferred. These issues are discussed in more detail below. We checked the characteristics of the data against those available in the relevant national censuses and, unsurprisingly, found some biases. As with much historical research the very poor are underrepresented because they are much harder to trace in the sources available, and the data contain more males than in the general population. This is because of the greater complexity of finding and linking sources for women who changed their name on marriage, and due to the fact that some respondents were interested in one-name studies and thus tended to trace their ancestors only down the male line. The data do cover most parts of the country, but with more information from the mid-nineteenth century when decennial censuses became available. Demographic bias is not restricted to the sex ratio of the data but also to age and marital status. The ancestors that respondents traced must have lived to be old enough to (in most cases) marry and have children, thus producing an ancestral line. However, we do not believe this to be a problem as there is no inherent reason why an individual who died as an infant or child would (had they lived) have had a different life trajectory to a sibling that survived into adulthood and had a family. Like all historical material there are inevitable biases and omissions in the data, but we are confident that this provides a more complete picture of life-time migration in Britain than any other available source.

The research produced many results which are fully reported elsewhere.Footnote ¹⁵ Here I focus on selected key issues and examine their significance for our wider understanding of migration history. First, the hypothesis proposed by Wilbur Zelinsky, of a mobility transition from low to high levels of mobility that moved in parallel with the demographic transition, has been influential within migration research.Footnote ¹⁶ However, the longitudinal data spanning some 200 years suggests that there was remarkably little change in levels of mobility in Britain. In all time periods most moves were over short distances, while both the frequency and characteristics of moves changed little over time. Second, the longitudinal data allow further assessment and refinement of Ravenstein's well-known ‘laws of migration’.Footnote ¹⁷ While Ravenstein's assessment that most moves were over short distances is clearly supported by the evidence, other assertions are, at best, only partially shown to apply. The British migration system was much more complex than the picture that Ravenstein painted of movement mainly up the urban hierarchy in a series of steps from small to larger places. In fact, the longitudinal data show that there was movement both from small to large settlements and from large to small settlements, with most moves in each direction of similar distances. Though this does support Ravenstein's (slightly contradictory) view that each flow had a counterflow, it fails to support his assertion that British towns grew largely by in-migration. As other studies have also shown,Footnote ¹⁸ in the majority of settlements natural increase was the more important demographic variable, albeit fuelled by migrants of child-bearing age. Ravenstein also suggested that men move further than women but that women move more often. The longitudinal data do not support this as both male and female mobility were broadly similar, in part due to the large volume of family migration that was shown to take place. Most of these discrepancies can be attributed to the sources most commonly used by British migration historians. Their reliance on decennial census and other static data, together with a focus on movement between relatively large areas, provides only a partial picture of the British migration system. By charting the migration history of a large number of individuals over their entire life courses a rather different picture emerges, one that is much more similar to the migration patterns and processes examined in much of continental Europe as revealed from the study of population registers.Footnote ¹⁹ The study of longitudinal migration data also allows individual migration experiences to be more firmly embedded in the social, economic and cultural landscape of the time. In the following section these data are used to illustrate some of the complexities of both internal and external migration.

4. Migration life histories

James (#4081)Footnote ²⁰ was born in 1812 in Yorkshire and as a baby moved with his parents from Aldbrough Beck to Middleham (North Yorkshire), a distance of about 30 km. He remained in the same location until age 29 when, now married and with a young family, he moved some 65 km to Durham city. In doing so he changed his work from agricultural employment to semi-skilled work in an iron foundry in Durham. Approximately ten years later he moved a short distance (14 km) to Crook in County Durham to a new employer but remained as an iron foundry worker. However, by 1861 he had moved 14 km to nearby Spennymoor (also in County Durham) to return to agricultural employment. This clearly did not work out as intended and in 1863 (age 51) he left his wife and family in Spennymoor and moved alone to Prussia to again work in an iron foundry. There is no direct information, but it can be reasonably assumed that this long overseas move was facilitated by a specific work opportunity, possibly with one of his County Durham employers expanding their operation to Prussia and sending workers there.Footnote ²¹ It seems unlikely that the move was speculative. His precise length of stay in Prussia is not known, but by 1871 he was back in Spennymoor, this time farming on his own account. Clearly his sojourn abroad had been sufficiently lucrative to enable him to acquire a farm in County Durham when he returned home. Work opportunities or needs clearly played a significant part in all of James' adult moves, and the relocation overseas was part of a process of moving between agricultural and industrial work with the aim of career advancement. I suggest that the move to Prussia was to some extent facilitated by previous moves within Britain, and that both internal and external migration should be viewed as part of a continuing process.

Richard (#2783) was born in Yorkshire in 1866 and age 13 moved with his parents from Sheffield to Hull. This move of some 100 km was the result of an occupational change by his father. In 1884 (age 18) Richard left home and moved over 200 km on his own to Wolverhampton (Staffordshire): the move was apparently precipitated by an unspecified family crisis, but we have no information on Richard's employment at this time. However, three years later and still unmarried he moved from Wolverhampton to Clapton (London) to take up an apprenticeship as a clerk, where he lodged with his master. He lived in Clapton for two years but then moved, again alone and unmarried, to Finland. He continued to work as a clerk, and it is a reasonable assumption that this move was facilitated by his London employer. We have no further record of him until 1906 when, age 40 and married, he moved from Denmark to Cambridge with his family, now in a more senior professional position. Back in Britain he and his family moved five more times before his death, with most moves apparently due to changes in employment. As with the case of James, multiple work-related moves in Britain and continental Europe were clearly linked together over the life course, with the experience of one feeding into the next and helping to build a career that spanned at least three countries.

One further example again illustrates the connected nature of internal and external moves and the ways in which together they made sense over the life course. David (#5810) was born in Yorkshire in 1824 but in 1836 (age 12) he moved alone from his family home in Leeds to Burslem (Staffordshire) to be apprenticed in the pottery industry of that town. No specific reason is given for this choice of employment, but it was to form the basis of his future career. In 1849, age 25 and now married, David moved just 1.5 km to the Cobridge area of the collection of settlements that form the Potteries, but four years later he made a much longer move to Spain where he continued to be employed as a skilled pottery worker. Specific reasons for this move are not clear: he may have been moved by his employer who could have had commercial interests in Spain, or he may have had more personal motives for this relocation to continental Europe. He remained in Spain for seven years, and in 1860 returned to England with his family, this time living in Bovey Tracey in Devon. Bovey Tracey had a thriving specialised pottery industry in the nineteenth century, and this is where David gained employment.Footnote ²² As in the other examples, the move overseas from Britain formed part of a number of internal and external moves that, I argue, fed into each other and formed part of a continuing process of personal and career advancement.

5. Migration motives

We know that migration motives are complex and that people move home for many different and often inter-connected reasons: work, marriage, family, housing or just whim.Footnote ²³ Precise reasons why a move takes place are usually hard to pinpoint from historical data unless there are very specific diary or letter sources that explain the move. Even then it may sometimes be the case that the reasons stated in (for instance) a letter to a relative may not be the only or underlying factor that stimulated a move.Footnote ²⁴ Data on the reasons for movement must thus be treated with caution. In my own research with Jean Turnbull on migration in Britain (outlined above) we sought to ascribe reasons for mobility in the past from the evidence provided by family historians, though motives were often inferred rather than directly specified.Footnote ²⁵ For the period 1840–1919 we calculated that 40–50 per cent of all moves could be principally attributed to the desire or need to seek new employment, with broadly similar proportions for both men and women, and across all age groups (apart from those over 60) and for most occupations (Table 1).

Table 1. Selected reasons for migration by key characteristics, Britain 1840–1919 (%)

Source: 16,091 migration life histories collected from family historians. See C. Pooley and J. Turnbull, Migration and mobility in Britain since the eighteenth century (London, 1998).

Percentages do not sum to 100% because only the principal inferred motives are included in the table.

If we focus specifically on migration from Britain overseas the relative significance of likely work-related movement varied considerably depending on the destination. For migrants from Britain to Western Europe approximately half of all such moves could be reasonably confidently described as labour migration – movement specifically for work-related purposes – but most of the rest were undertaken either for education or the rather vague term of emigration. It can be suggested that most moves for education were designed to improve career prospects, and that although the decision to emigrate could be stimulated by a variety of factors work was likely to be one significant dimension. Work-related moves also dominated in migration from Britain to Ireland and to Asia, but for movement to the USA, Canada, Australia and New Zealand the more general motive of emigration accounted for at least three quarters of moves. Employment prospects in some form were clearly uppermost in the minds of most migrants who left Britain for other countries in the past.Footnote ²⁶ However, it was also evident from the data that such categories were by no means clear cut, and that for many people moves that could be classified in one way (such as a desire for new employment) were also influenced by a range of other factors. The complex mix of factors that could stimulate migration may be overlooked if studies seek to ascribe a single cause, such as labour migration or family migration, to a move.

6. Using life writing to understand migration motives

It is hard to identify motives for residential migration even in contemporary populations as respondents to interviews or surveys may reveal only some of the reasons why a decision to move was made and, in particular, may conceal negative factors (such as poverty) and accentuate more positive reasons such as wanting a more attractive location. Identifying motives for migration in the past is even more difficult. Although for the recent past oral testimonies may be used, such data obviously rely on memory and provide even more scope for post-migration rationalisation of motives for a move.Footnote ²⁷ For the more distant past life writing offers the most productive avenue for identifying migration motives, but there are many limitations to what such sources can reveal. The term life writing can cover a number of different forms including personal letters, diaries, autobiographies, life histories and even some creative writing that draws on an author's own experiences.Footnote ²⁸ I have been using such documents to study aspects of mobility and migration in Britain from the 1840s to the 1950s,Footnote ²⁹ and I draw on a selection of these to illustrate the complex combination of factors that can generate residential mobility.

All life writing will be biased towards those with the time, literacy and motivation to record aspects of their life; the poorest are rarely directly represented in such accounts. We also have no way of knowing what a writer chose to include and what they omitted, either deliberately or by chance. Likewise, survival of such documents is largely random, so it is not possible to assess representativeness. Life writing sources must be seen as individual examples that may be used to illustrate an issue, but it is not possible to generalise from such sources to a larger population. Personal diaries, written up daily (or at least frequently) are likely to be the most useful and reliable. They should provide an almost immediate record of events and may thus record feelings that occurred at the time. While diaries produced by the rich and famous may have been constructed with a view to later publication and were thus carefully constructed to convey a particular image, this was much less likely to be the case for diaries written by the majority of ordinary people who lived unremarkable lives. In this research I have avoided using diaries that were obviously destined for later publication. Although diaries do survive for men and women of all ages, young women appear to have been the most prolific diarists among the general population creating a further potential bias in the data. It can also be argued that diaries are most likely to record events that were unusual or remarkable rather than the mundane or everyday. This can seriously skew our perception of the past as those things that were most commonplace (such as everyday mobility) become erased from the record. However, despite such limitations, diaries can provide important insights into both daily mobility and residential migration when it occurred.Footnote ³⁰

Collections of letters written by the general population occur much less frequently than diaries and are limited in that they occur only sporadically (when someone chose to write) and, usually, consist only of a one-way conversation. However, where they do occur, they can provide detailed accounts of the actions and motives of the writer in relation to specific events such as migration. However, as with all documents, migrants' letters must be interpreted with caution. Migrants may have deliberately painted a positive view of their experiences for their readers back home, and collections of emigrant letters are obviously only representative of a small proportion of the population.Footnote ³¹

Autobiographies and life histories occur rather more frequently than diaries, and many have been published and are thus easier to access and read than hand-written, unpublished diaries deposited in record offices around the country. They also usually cover most of a life span, thus encompassing almost all residential moves, whereas surviving diaries often cover only a short portion of a lifetime. However, I argue that autobiographies and life histories are more limited in their usefulness than diaries in that they almost inevitably present a carefully considered narrative of the past that may be designed to project a particular image or to justify past actions. Such documents were much more often written by men than by women, usually later in life when events could be viewed and reinterpreted with hindsight. Although major life events such as migration are likely to be recorded, more mundane activities are much less likely to have been included than in diaries.Footnote ³² In this paper I use a mixture of diaries, letters and life historiesFootnote ³³ to illustrate different aspects of the migration decision-making process, but all the above caveats must be borne in mind in interpreting the evidence. Brief information about the writers and sources is given in Table 2, with further elaboration below.

Table 2. Life writing sources used in the paper

7. Deducing migration motives from life writing

The life history of Benjamin Shaw,Footnote ³⁴ apparently written up from diaries he had kept (but which do not survive) recorded many instances of migration for various different reasons. Shaw was born in 1772 in a hamlet close to the village of Dent (then in the West Riding of Yorkshire, now in Cumbria), and wrote his life history in 1826 while living in Preston (Lancashire). He recorded details of his life and that of many of his family members, and when analysing the life history we were able to check many details against other sources to corroborate the information. One event in particular illustrates the complex family-related decisions that led to a move that might on the face of it be assumed to be for employment. In the late-eighteenth century Joseph Shaw (Ben's father) and his family had been working as handloom weavers in Dent. However, the trade was declining, Joseph's wife Isabella was in poor health and could contribute little to the domestic economy, and they were finding it increasingly hard to make a living. In 1791 a mill owner in Dolphinholme (Lancashire) was recruiting labour in the Dent area and Joseph (together with some other families from the community) decided to move with his wife and seven children to Dolphinhome to work in the mill. Benjamin wrote in his account:Footnote ³⁵

If seen from the perspective of static sources such as the decennial census it is likely that a move such as this would be identified as work-related. Although it did benefit the family's income, the move was not of direct benefit to the principal wage earner and head of household. Rather, Joseph Shaw's status and earning power was almost certainly eventually diminished by the move, but the opportunities for his children were greatly increased. Although Joseph was initially taken on in the machine shop at the mill, and his children given apprenticeships, within three weeks he had argued with his employer over wages and had been sacked. Although later taken on again at Dolphinholme this was only piecing rovings, work often done by children, and was most likely only accepted in loyalty to his family.Footnote ³⁷ The alternative would have been to leave them in Dolphinholme and seek work elsewhere. The move was thus designed to benefit Joseph's children and to increase family income and security rather than simply improving the employment potential of the main decision maker, which it did not achieve. In practice this was a migration decision taken for a complex set of motives including work, family and health, though superficially it could easily be labelled as labour migration.

Female employment in the nineteenth century was more limited than for men,Footnote ³⁸ and most married women would necessarily move with their husband and family, but the migration decisions of single women could also be complicated. The letter books of Ellen WeetonFootnote ³⁹ provide one example of a single woman who appeared to move home primarily for work-related reasons, but with other underlying factors also affecting the decision. Ellen (Nelly) was born in Lancaster in 1776, but her father died when she was almost six years old and her mother then moved to Up Holland (Lancashire) in 1784 to establish a small village school. However, in 1797 Nelly's mother died and at the age of 21 Nelly was left alone to run the school. She also discovered that her mother had accumulated substantial debts that she was now responsible for. Nelly remained at the school in Up Holland for 11 years and succeeded in paying off the debts she had inherited, but it seems that the place had bad memories for her and once free of debt she was determined to leave. Her aim was to seek a post as a teacher or governess, preferably in an urban location.Footnote ⁴⁰ In a letter dated February 1808 Nelly wrote:

Some two months later all arrangements had been made and, after a short stay with relatives, Nelly intended to stay with a friend in Liverpool while looking for more permanent lodgings and work on Merseyside. She wrote to her friend in Liverpool:

Nelly Weeton changed her work but her motives for moving seemed complex. They were partly generated by a simple desire to break the link with her mother's home (though she retained the house as an investment), partly to move to a larger community, and partly to find new employment. The idea of change seemed as strong a motive as any particular employment opportunity. She had no work to go to in Liverpool and planned to live on savings until something suitable came up. Although in one sense this was labour migration – she changed her work – there were a range of interlocked motives in play.

John Lee (born 1842) was a young man living in Heywood (Lancashire) when he decided to move to Merseyside in 1861/62. While this decision seemed to be based primarily on the desire to improve his employment prospects and income, there were also other factors involved. John was employed in a junior capacity in a retail drapery business. Although only 18 years of age he was keen to progress and to eventually set up his own business. He briefly considered an opportunity in the wholesale trade but was warned off this by more experienced friends as the trade was being badly affected by the onset of the American Civil War in 1861.Footnote ⁴³ In March 1861 he recorded a detailed conversation he had with his employer, ending by stating that:

By September 1861 he had made the decision to stay in his present employment over the winter, though by December he was again restless and had begun actively making contacts to try to find a new situation:

Eventually John Lee took the decision to leave Heywood in March 1862, moving to Liverpool to lodge with his aunt. He did not have a position to go to but did have some potential contacts. He made several enquiries over the following four weeks, eventually securing a position as a shop assistant in a drapery business in Birkenhead. His gamble in moving speculatively had paid off and, indeed, he did eventually progress to running his own business. He changed employers in Birkenhead in December 1861 and eventually became a partner in the business. He described his initial move and search for employment as follows:

Thus, although John Lee's move to Liverpool was undertaken with a view to finding new work and advancing his career, the choice of Merseyside as a destination was due mainly to the fact that he had existing contacts there and that he could, in the first instance, lodge with his aunt in Liverpool. These family connections were clearly crucial in convincing him that such a move was possible and led him to move speculatively before any new position had been secured.

Another example of mid-nineteenth century migration by a young man demonstrates even more clearly the way in which family ties and loyalties can influence a residential move that at first sight may be seen solely as labour migration designed to improve career prospects. Amos Kniveton was born in 1835 in the village of Astley Green, some ten km from Manchester.Footnote ⁵⁰ His father was a boot and shoemaker in the village and Amos was the second son in a large family of 14 children. After periods in a local cotton mill (where he started half-time at the age of seven) and as a ploughboy, at the age of 14 he returned home and was apprenticed to his father's trade. In 1857 he married and two years later established his own boot and shoemaking business in Astley, but in 1863 he applied for and secured employment as a ‘bootmaker and manager’ for a substantial business in Morecambe, north Lancashire. This took him some 70 km from his family home. On the face of it this move could easily be construed only as migration to secure more income and experience. However, although it did provide these advantages, in his life history Amos suggests a rather different motive for the decision to leave his home village of Astley: ‘As trade generally was good, my business venture proved a success, but finding it was to my Father injuring his business, I determined on the first opportunity to leave the place to himself’.Footnote ⁵¹ Amos prospered in Morecambe, moving on to a better position in Blackburn before in 1867 setting up his own business in Leigh just four km from his home. Thus the move did further his career, but it is clear that had it not been for the fact that he did not wish to compete directly with his father, and in doing so damage his father's trade, he would probably not have made the move to Morecambe at that time at least.

For young men and women leaving home for the first time to take up employment the decision was often not one over which they had much control. Parents or guardians would most likely identify a suitable trade and place for an apprenticeship or position of some kind. Although effectively labour migration – moving to take up work – the decision to move was taken by someone who did not move. Henry Jaques was born in Islington, London in 1842.Footnote ⁵² His father was a watchmaker and initially Henry was set to work helping his father. However, he showed no aptitude for or enjoyment of this work so in 1858 (at the relatively old age for an apprenticeship of 16) he was apprenticed to a drapery business in the locality.Footnote ⁵³ Although Henry was glad to escape from the watchmaker's workshop it is clear from his account that he had little say in precisely where he went or what trade he entered:

After one month Henry was deemed satisfactory and was indentured for three years with no salary but 2/- per week pocket money in lieu of a premium. Although he was dismissed from this apprenticeship after only 17 months this first employment away from home did enable him to develop skills that formed the basis of his later career. Migration for young men and women entering employment away from home for the first time was often something over which they had only limited agency, and the aspirations and contacts of family were often crucial in securing a position which may or may not have been to the satisfaction of the young person themselves.

Not all moves are undertaken voluntarily as some may be stimulated by hardship and circumstances beyond the control of an individual. William Holt's autobiography provides several examples of residential migration by his family, with one incident in the early twentieth century illustrating clearly the way in which the interaction of unfortunate events could lead to a forced change of employment.Footnote ⁵⁵ In 1906 William's parents had moved from Hebden Bridge (Yorkshire) to the nearby village of Charlestown. In Hebden Bridge Mr Holt had worked as a coal dealer but in Charlestown his firm had acquired an old brewery with the intention of making a better year-round income selling ‘stone pop, hop-ale, and dandelion and burdock stout’.Footnote ⁵⁶ William was only nine at the time so his memory may be partial, but according to his account this enterprise failed rapidly, and sometime between 1906 and 1911 (a precise date is not given) the family were forced by poverty and debt to flit to Todmorden where Mr Holt returned to coal hawking:

Once again, from this example, although the move could be characterised as one generated by the search for new employment, this was not sought voluntarily, and was the consequence of a series of economic and family circumstances, and probably unwise decisions, that eventually made a move inevitable.Footnote ⁵⁸

8. Concluding remarks

Most research in migration history places more emphasis on long-distance international migration than on shorter-distance internal moves, and usually focuses only on a small number of specific moves, or particular sectors of the population, over a relatively short period of the life course.Footnote ⁵⁹ Studies of emigration from Europe to North America are especially prominent, together with migration from Britain to Australia and New Zealand.Footnote ⁶⁰ Studies of internal migration in Britain have mostly utilised census data and have focused on movement between relatively large spatial units such as English counties. This has emphasised long-distance moves over short distance migration, and has also tended to overstate the extent of rural to urban movement.Footnote ⁶¹

The main purpose of this paper has been to demonstrate the ways in which a consideration of migration over a longer period of the life course, using a variety of sources, can illuminate the complexity of the migration process. In particular, I first stress the ways in which both internal and international migration may be linked together through an evolving personal history in which individual, familial and employment opportunities were changed and, in many cases, enhanced. Second, I use a range of life writing to examine the complexities of migration decision making. Although the end result of a move may have been new (and in some cases improved) employment prospects, the motivations for that move could have been generated by a complex set of personal and family circumstances. Indeed, the precise motivations for a move may never have been completely clear to the migrant at the time, though I suggest that few moves were completely speculative. It is only by using a range of life writing that such complexities can begin to be revealed. The argument that migration decision-making is complex is not novel. But demonstrating such complexities cannot readily be achieved with most aggregate sources that are commonly used to study international migration. The problems inherent in the categorization of migration as (for instance) labour migration or family migration, and the reasons why such labels are commonly applied, are clearly explained in a recent essay by Marlou Schrover.Footnote ⁶² A focus on life writing and longitudinal life histories enables the complexities of individual migration trajectories to be more effectively explored, although as with all historical sources they can never provide answers to all the questions that might be asked.