Tsoukalas sees out-migration from the countryside as a process of upward social mobility, the children of the village farmer becoming petit bourgeois in the urban centre, though not necessarily a Greek urban centre (Tsoukalas, 1987, 136-46). The single most important “weapon” of those migrants in achieving such social mobility was education, which was provided freely and rather equitably by the Greek State from an early date (Tsoukalas, 1987, 517-26).

Little is known about the characteristics of the migratory moves, whether they were permanent or temporary, seasonal or circular. In many studies of Greek migration, the type of movement is mentioned in passing. Interestingly, there is a significant number of such passing references to circular/seasonal movements of Greek populations stretching from the eighteenth to twentieth centuries, and to a variety of destinations (Paspatis, 1862 in Liakos 1988, p. 45; Mpafounes, 1984, 563; Gounaris, 1989, 134 et 149; Campbell, 1979, 12-3; Caftanzoglou, 1997, 412-3; Frangakis-Syrett and Wagstaff, 1996, 129 et 133; Psuchogios, 1985, 27; Tsoukalas, 1987, 150-1; Gage, 1989, 443; Arnold Costa, 1988, 180). (Among the above, only Arnold Costa utilised the term of circulation).

In this paper, the Tsoukalas’s propositions are tested and some of the questions raised above are answered by an in-depth study of the island population of Mykonos. This paper is divided in four parts. First, a brief background to Mykonos is presented and the data utilised in this study are outlined. The combination of quantitative and qualitative data provides a rounded and unexpected picture of migration patterns. The demographic background of the island population is examined in the second part, revealing very high levels of out-migration. In the following section, the main destinations of the migrants and their migration patterns are presented, emphasizing the dominance and continuity of circular migration among the members of the working class throughout the century and irrespective of migration destination. The importance of sex and life cycle stages is stressed when the focus on migration turns to Athens. In contrast, wealthy migrants were in most cases engaged in long-term migration. In the final part, the main findings of the paper are discussed and the possibility of the dominance of circular migration among other Greek populations is raised. The conclusion is reached that migration, rather than being a strategy of upward social mobility as Tsoukalas claimed, it was a strategy of retaining social class status, a way to avoid downward social class mobility. Circulation averted to a large degree the creation of a “permanent” proletariat within Greece up to 1920 by removing—usually temporarily—the landless workers to foreign labor markets.

There are two main settlements, Mykonos town, where the island port is located, and the inland community of Ano Mera. The latter consisted up to the Second World War of scattered farmsteads, far apart from one another. Farmers dominated Ano Mera whereas in the town those with a monetary income were established, namely artisans, shop-owners, families of sailors, civil servants, the upper class but also any landless workers. Mykonos is characterised by sparse, rather infertile soil. The island economy was highly monetarized as extremely frequent purchases and sales of land and houses from the middle of the eighteenth century demonstrates (Stott, 1982; Demetropoulos, 1997). On Mykonos, as in the rest of the Cyclades, small-holding prevailed.

The island of Mykonos was selected for this study for a variety of reasons. First, it is typical of small Cycladic islands and to an extent it is typical of Greek small islands in general. This statement holds up to the end of the Second World War, a point that signalled the diverse economic evolution of different islands depending on the onset and level of tourism. Second, Mykonos has been part of the Greek State since its creation in 1828, unlike many other Greek communities. Moreover, Mykonos is a rare case of an island where a variety of data exist and where these sets of data have been employed in a number of studies of its society and population (Stott, 1982; Hionidou, 1993; Demetropoulos, 1997) [3]. Many of these sources are used in this study. The civil registration data of the Mykonos population, available from 1859, are employed in summarising the demographic background of the population. (For a discussion on the quality of these data see Hionidou, 1993, 21-31; Hionidou, 1995b, 127.) The method of family reconstitution enabled the cons-truction of sophisticated measures of fertility, infant and child mortality as well as nuptiality (Hionidou, 1993). Also, the 1861 census manuscript of the Mykonos population allows me to trace the pre-sence of in-migrants on the island. The 1861, 1870 and 1879 partial census manuscripts of Hermoupolis, Syros are used to examine migration from Mykonos to Hermoupolis. A list of conscripts residing outside Mykonos at various years between 1860 and 1880 traced in the archive of the folklore museum of Mykonos is also used.

Lastly, and most important, the 28 interviews conducted on Mykonos in March and April 1994 with elderly Mykoniati provide the focal point of this paper. These interviews were part of a wider study of the demographic beha-viour of the Mykoniati including their contraceptive knowledge and practices (Hionidou, 1998). Hence twenty-five women and seven men were interviewed, chosen through personal contacts (three couples and two sisters are included in the above figures). While no effort was made to select a “representative sample”, overall, the informants seemed to cover almost the whole spectrum of the possible alternatives of place of residence (current and past), migratory moves, marital status, occupation and social status. The ages of the informants ranged from 70 to 93 with a median age of 84.5. A set of basic questions was devised. Each informant was asked to present her/his own life-path, including migratory moves and employment history. Moreover, they were asked to present the life-path of each of their parents, each of their siblings as well as those of their own children. In addition, they were explicitly asked if they were aware of any other relatives that had migrated to the specific destinations known to have attracted Mykoniati over time or anywhere else. The above process ensured that the focus of the answers was on the individual life-course rather than migratory movements and thus, to a large degree, the problem of “selective recall” was avoided. The fact that the answers provided were only age-specific i.e. persons of the same age pointed to the same migratory destinations irrespective of their sex and socio-economic background suggests that such a bias was overwhelmingly avoided. Naturally, some marginal migration destinations may have been omitted but this would have not affect the overall picture.

The interviews were open-ended and the informants' answers were followed up. This enabled to receive their views and understanding of the processes involved rather than imposing my own preconceived expectations on them. It is significant that these informants were persons who eventually resided on Mykonos and thus, in a broad sense, they constitute the “stayers” rather than the “movers”. Still, only eight of my informants were living at the time of the interview in the same locality where they had been born without ever having moved away.

Before turning into the main findings, a demographic background of Mykonos is required in order to place the estimates of the volume of migration from the island into a wider perspective.

Two distinct eras emerge, either side of 1910. Initially, reproductive levels were high, with the total marital fertility rate (TMFR) ranging from 9.2 for those couples married in 1889-98 to 10.4 for those married in 1899-1908. Fertility control was introduced first by those couples married after 1914, meaning that fertility was declining by the late 1920s (Hionidou, 1998, 71). Mortality was rather low, with an estimated life expectancy at birth in 1861 of 47.4 for males and 43.7 for females (Hionidou, 1993, Table 6.7). Levels of Infant mortality were also rather low, never exceeding 110 before 1919-28, when a temporary increase in mortality was observed (Hionidou, 1997, fig.1). Infant mortality decline occurred essentially in the 1930s, the decade of important demographic changes. Up to that date, a significant population surplus was created—approximately two per cent per year—which is not reflected in a population size that remained virtually stable at around 4,500 persons. The answer to the extensive natural population growth was migration. As we will see later, migration into the island was insignificant and therefore the net migration rates reflect levels of out-migration. Out-migration was responsible for “removing” from the island the number of persons almost equal to those “removed” by death [4]. After 1930, the rates of natural population growth were considerably lower, less than one per cent per annum, mainly due to the declining fertility. Still, out-migration rates remained at the earlier levels, except for 1961 when they declined. This decline may be associated with the beginnings of extensive tourism that increased the employment opportunities on the island (Stott, 1985, 196-9). The population size declined in the 1940’s partly as a result of the 1941-42 famine and remained at this reduced level until 1961 (Hionidou, 1995c).

Clearly, migration was a very significant variable in the island’s demographic profile and, as we will see later, part of people’s every day lives. The extent and character of in-migration is the focus of the next section.

Tab. 1
Population enumerated on Mykonos in 1861, by sex and place of birth

A noticeable aspect of the in-migrants' profile in 1861 is that they held some of the most prestigious occupations on the island. The only doctor, most of the civil servants, the midwife, some of the teachers, and 41 per cent of merchants were non-Mykoniati (1861 Census manuscript, Municipal Library of Mykonos). This points to the high mobility of the “educated” and the affluent within Greece, irrespective of their place of origin.

Seafarers dominated the 1861 in-migrants, though their presence was reduced over-time with the declining importance of sea faring in general and that of the Mykonos port in particular. So, sailors marrying on Mykonos constituted 45 per cent of all non-Mykoniati grooms in the period 1859-1868. This declined to 32 per cent in the following decade and further to 15 per cent in 1889-1898, never to rise above 10 per cent thereafter. The existence of only six households in 1861 with more than one in-migrant member suggests that the in-migrants, or at least those who remained long enough on the island, were not isolated from the native population. Rather, in most instances the long-term in-migrants were cases of inter-marriage.

The presence of refugees from Asia Minor who arrived in 1923 is not appa-rent in Table 2, despite the fact that a number of marriages involving them took place mainly in 1919-1928.

Tab. 2
Brides and grooms born outside Mykonos as a percentage of all marriages

These refugees were “allocated” to Mykonos by the central government after their arrival in Athens in 1922. In 1924 Kyriazopoulos reports the existence of approximately 30 refugee families on the island, although by the end of the decade the majority of them had left (Kyriazopoulos, 1924; Stott, 1982, 56).

Finally, the places of origin of the in-migrants as observed in the marriage records were primarily from within Greece, with the obvious exception of the Asia Minor refugees. Interestingly, the overwhelming majority of in-migrants originated from islands and coastal towns, the obvious reason being the importance of sailing for communications and for creating links between those communities [7].

In summary, long-term in-migration did not attain significant levels on Mykonos between the mid-nineteenth and mid-twentieth centuries. Still, it was much more important for men than for women. Now, we turn to out-migration, investigating its levels, destinations and patterns.

Tab. 3
Mykoniati population enumerated on Mykonos, Greece and abroad

In Greek censuses, apart from the de facto population of a community, the number of the members of the community enumerated outside that community and within Greece is also cited. In the overwhelming majority of cases the membership of a community (demoti-kotes) refers to place of birth, which, in addition, is the place where a person exercises his/her civil rights. Nevertheless, the membership of a community can change during someone's lifetime. For example, women often transfer their membership to the community of their husband, although this is not always the case. Children become members of the community in which their father is a member, even if they were born elsewhere. Accordingly, the figures in Table 3 should be treated with caution. Finally, the figures referring to those enumerated abroad were provided by Greek consuls. The census organisers repeatedly admitted that the numbers provided grossly under-estimated the extent of emigration (1907 Greek Census). Table 3 indicates that the numbers of Mykoniati, or more correctly Mykoniati and their descendants, living away from Mykonos occasionally approached those living on the island. While the approximate nature of these data should be stressed, the figures suggest very significant levels of out-migration, possibly rising at the beginning of the twentieth century, with both international and internal destinations. Emerging both from the archival and the oral accounts is a rather impressive list of destinations and a distinction in the migratory patterns before and after 1910.

For Mykonos, a broad picture of out-migration destinations emerges from lists of civilians who were abroad at the time of their conscription in years between 1860 and 1880 (Table 4).

Tab. 4
List of conscripts (klerotos) residing outside Mykonos at the time of call-up (year of call-up and place of residence)

For most of these conscripts, a place of residence is given. Where this is not given (unknown/irrelevant) there is either an indication that the conscript was deceased or presumably was a sailor. In descending order, the main destinations of young, single males were the Lower Danube/Romania, the neighbouring island of Syros, Russia and to a lesser extent Piraeus/Athens, the Ottoman Empire and Egypt. The oral accounts and the secondary sources also refer to these destinations (Yiangakes, 1985; Evangelides, 1912; Stefanos, 1884, 555).

The exclusive choice of coastal areas as destinations is only too clear when the specific destinations are considered, and in most cases this choice was associated with extensive links that had already been established through sailing and trade: “Whoever was not a seafarer, would be working on the River [Danube] or in various ports of Russia, where from they would return to the island [of Mykonos] at specific times of the year [during festivals]” (Mykoniatika Chronika, 19 August 1934 referring to the 1880s). This commentator himself was born in 1867 on Mykonos, son and grandson of captains and owners of sai-ling ships. He spent a large part of his life working in Verdianska, Russia as a merchant. Later, he worked in Piraeus and he eventually retired on Mykonos (Kousathanas, 1986, 387-8). A prospe-rous community of Mykoniati migrants existed in Verdianska but also elsewhere in Southern Russia during the second half of the nineteenth century and until the Russian revolution (Stefanos, 1884, 555). Among the 28 interviews there are four references to migrants residing in Russia, two of which concern families where the father was living and working there in connection with the wheat trade, while the rest of the family was circulating between Russia and Mykonos according to its life-cycle stage. (So, for example, when the children were young the family resided in Russia. When the children became of school age the family moved back to Mykonos while the father remained in Russia (No 18 and No 25).) Russia was a place where regular wage employment was readily available, contrasting the situation on Mykonos: “They [Mykoniati men] were in great demand there. It was not a matter of concern whether they would find a job or not.” (No 25, male, born in 1905). The extent of migration from Mykonos to Russia was very significant: “Half of Mykonos was there then… Russia would feed all the people then.” (No 25).

The origins of this chain migration can be sought among the establishment at the Russian ports of some affluent Mykoniati such as Morfinos who owned a wine-making factory, and Ampanopoulos, a grain trader in Verdianska. Their pre-sence ensured employment for any newly arrived Mykoniati, at least for the initial stages of this migratory move (Kousathanas, 1986, 390-1; No 18).

Thus, for the less affluent Mykoniati, migration was triggered by the availability of cash employment and the widespread knowledge of that availability. Moreover, they were encouraged by the existence of a Mykoniati nucleus and utilised that to find employment [8]. A similar situation seems to have existed in Odessa, another Russian port and migration destination of Greeks. Affluent Greeks had long been established in Odessa (Stavrianos, 1958, 481; Herlihy, 1989, 239-40). Herlihy found that 27 per cent of Greek migrants were living in the most affluent area of the city in the last decade of the nineteenth century, while another quarter was residing in some of the poorest areas (Herlihy 1989, 238-9). While many poor Greeks were arriving in Odessa in the 1880s, the more wealthy Greeks had not only stopped migrating there, but also some of those already established were leaving in search for opportunities elsewhere (Herlihy, 1989, 241-2). Thus, the timing of migration was highly wealth-specific, with the wealthiest having migrated earlier and the less wealthy having followed some decades later. Psuroukes suggests that only in the1860s did members of the Greek working class begun to appear in the Paroikies (Greek emigrant communities) where for centuries well-off Greeks were established (1983, 140). Moreover, it seems that the movement of the affluent was long-term whereas that of the poor was circulatory/temporary, aiming at the accumulation of capital and eventual return to the island [9].

The Greek presence in the Lower Danube goes well back into the eighteenth century. By the nineteenth century, their presence in both navigation and trade had become predominant (Focas, 1988, 119-20). As in Russia, the Greeks of the Danube were of a mixed socio-economic background (Focas, 1988, 122). In the interviews conducted among the elderly Mykoniati, there are two references to Mykoniati travelling/residing in Romania. These concern the sister of an informant who got married to a Mykoniati who already lived in Romania but subsequently moved to Piraeus (No 5); and the father of an informant who spent his working life there while his family resided on the island (No 22). Both instances refer to the first two decades of the twentieth century. The revival of nationalism in the Balkans and the subsequent restrictions imposed on potential in-migrants among the Balkan countries brought to an end migration to Romania (Fairchild, 1911, 74-5). At the same time Russia also started restricting the entrance of Greeks and the Russian revolution of 1917 brought it to a definite end. There are also some references indicating the migration of Mykoniati to Egypt but this does not seem to have been extensive (No 1; No 24).

The reason for the small number of references in the oral accounts referring to the above destinations is simply that only the oldest of the informants had a primary or secondary experience of such migratory moves. Another equally important reason for the lack of refe-rences is that after the 1917 Revolution many of the migrants to Russia returned initially to Mykonos. Once in a familiar and supportive environment, they sought to arrange employment through existing channels of migrants established elsewhere and subsequently moved on to wherever they could get employment (mostly Athens and Piraeus) (No 25). Only those who were unable to obtain employment elsewhere or were too old remained on the island. In other cases, a direct movement to Piraeus or elsewhere most probably took place, thus restric-ting the possibility of obtaining further references from the informants.

In respect to migration within Greece, movement to Syros and to Athens/ Piraeus was also taking place. The proximity of Syros to Mykonos, in addition to the continued importance of its port long after the decline of that of Mykonos, attracted some Mykoniati migrants (Table 4). Mykoniati seafarers and wor-kers were looking for employment opportunities in Syros throughout the nineteenth century (Yiangakes, 1985). The census returns of Hermoupolis, the main town on Syros, reveal the extent of Mykoniati presence there, as well as their employment and status. The presence of Mykoniati on Syros almost doubled from 1861 to 1879, constituting 11 per cent of the Mykonos population in 1879 [10]. They were overwhelmingly “families”, with Mykoniati servants in many cases, with the bread-winner either being merchant/Captain or labourer (not agricultural), thus indicating the migration of both well-off and working class Mykoniati to Hermoupolis (No 27). Among the oral accounts there was a reference to a Mykoniati girl marrying to a Mykoniati already residing in Hermoupolis who established their household there. Moreover, in many cases it was mentioned that boys were sent to Syros as servants/ apprentices (paragios) and girls were working there as servants (No15, No 20, No 24) [11].

During this period, that is prior to 1910, Athens and Piraeus offered very few opportunities for paid employment to labourers. The growth of both cities during the nineteenth century was mainly based on the arrival of refugee populations, but also on in-migrants who were overwhelmingly engaged in domestic service (Leontidou, 1989, 63). The few references to migration to Athens from our informants mention single Mykoniati girls working as servants or boys working as apprentices [12]. Also, at the end of this period, the fathers of three informants would seek temporary wage employment in Athens while their families resided on Mykonos. One of these temporary migrants was employed in quarries, as a builder and a carter (No 15), while the other two were employed in the port of Piraeus as labourers (No 13, No 14). As will be seen, they were among the first to be followed by many Mykoniati on a circulatory movement between Mykonos and Athens that cha-racterised the years to follow.

The situation was different for the wealthy Mykoniati. By the end of the nineteenth century the elite of Mykoniati society took up permanent residence in Athens, since their economic interests had already been established in the capital (Stott, 1982, 206). Joining them were the wealthy and educated Mykoniati who had to leave Russia and Romania at the beginning of the twen-tieth century. The permanent movement to Athens during this period seems to have been restricted to affluent families, while the majority of the movements of the working classes to Athens/Piraeus were of a temporary and/or circulatory nature.

Migration from Greece to USA started in the 1890s and by 1907 its extent began to startle the Greek government (Saloutos, 1973, 409; Fairchild, 1911, 227-8). Nevertheless, it never acquired similar proportions among the Mykoniati. The first out-migrants to the USA left Mykonos in 1909-1913 to establish themselves in Joliette, Illinois (Yiangakes, 1985; oral accounts). This movement was not considered permanent during the early stages. Two of the informants explained how their fathers went to Joliette while married, leaving both their wives and children on Mykonos. In each case, the migrants left in the early years of their marriages in order to amass savings that were, in most cases, invested on the island where their families remained. One of the two migrants died in Joliette in 1913 in an occupational accident (No 16). The other returned two years after his departure because he was conscripted (No 11). Both these migrants were working in construction in USA and they did not have any relatives in Joliette, but “there were other Mykoniati there” (No 11). Five more informants mentioned that members of their families emigrated to the USA. In all these cases some relatives were already in Joliette to help them and the migratory movement was clearly permanent. These references stretch from 1915 to 1952 and it seems that when members of a family were involved in permanent emigration to USA, inva-riably more than one made the move. (As members of the family I have included spouses, children, siblings, nieces and nephews). This emergent pattern was a result of the restrictions imposed by the USA government in the early 1920s. The Immigration Quota Act of 1924 diminished the number of Greeks who could enter USA in any year (Saloutos, 1973, 424-5). Thus, those who could hope to emigrate to USA were exactly those with relatives already established there. This explains the high concentration of emigrants to USA in specific families.

Without exception, all of the Mykoniati emigrants to USA went to Joliette, Illinois. This chain migration is vividly described by one of the informants whose four siblings all left:

A. […] One day my father came home and told mother to write a letter to her brother K [who was in Joliette] so that D [the informant's eldest brother] could leave the island. “I am not having my son becoming a fisherman like the others” he said.

Q. And where did he go?

A. He went to the USA where my mother's brother was. My father arranged for his passport and [my brother] went.

Q. When did he go?

A. […] D went in 1914-1915. To Joliette. […] He was 15, 16 years old.

Q. And what did he do there?

A. A first cousin of ours had a bakery there. So, he took D with him…

Q. Did your brother stay in Joliette?

A. Yes, all his life. He died there.

Q. And your sister M?

A. M was really close to that brother of mine… So, D said he would take her over [to the USA]. She was single… My mother's brother was there also. She left 18 years old… And my younger brother wanted to leave also. He left 14 years old. He left with my sister…

Q. What did M do? Did she get married?

A. She was in Joliette. There K [a Mykoniati] —Joliette was like Mykonos— men-tioned that a Greek man wanted to marry a Greek girl.

Q. Was he Mykoniati as well?

A. No, he was from S. [mainland Greece]. So, he came to Joliette. He was a well-established young man. So, they got married and she went to Oklahoma [where the groom lived]…

Q. Did your younger brother get married?

A. Yes, he got married to a Mykoniati girl from Joliette…

Q. So, you were the only one who stayed on Mykonos?

A. My sister L was also here. Her husband was in USA. […] She got married in 1930 and then he went back to USA.

Q. And when did they stay together?

A. […] He came back four years later and left again. […] She left with her children after the Second World War [to the USA]. [No 18]

While similar levels of emigration were not observed for all the families with a link to the USA, this example illustrates that those who had that option made liberal use of it. Moreover, the above informant illustrates the circulatory pattern of migration to USA of married men—or single ones who returned to Greece to get married—whose family remained in Greece. (The return of the circular migrants would usually last a number of months and only rarely more than a year).

In the immediate post-Second World War period some further migration to USA was allowed and a small number of Mykoniati emigrated once again to Jo-liette (Yiangakes, 1985; No 3). Approximately 200 families—all descendants of Mykoniati—were living in Joliette in the 1980s (Yiangakes 1985). So, while emigration to USA started only sometime in the twentieth century with both temporary and permanent movements, it became clearly permanent once entrance restrictions were imposed. These restrictions, and the fact that Mykoniati were late-comers in this migratory movement, also meant that emigration to USA from Mykonos did not reach the very high levels observed elsewhere in Greece.

By the early 1920s all the old destinations of Mykoniati were inaccessible with one exception, Athens/Piraeus. (Here a distinction between the two cities is not made if only because of their physical proximity.) The economic expansion and the industrial development there during the inter-war period meant that opportunities for wage employment were better than ever before (Leontidou, 1989, 166-79). This was the almost exclusive destination of Mykoniati migrants for the rest of the twentieth century. The emerging pattern of migratory movements to Athens were a continuation of some of the earlier ones such as young single women working in service; young single men looking for seasonal cash employment; married men circulating between Mykonos and Athens while their family resided on Mykonos. Some new trends also emerged, such as the “permanent” migration of married couples. (Many returned to Mykonos upon retirement.) These were the characteristic patterns of out-migration from Mykonos to Athens prior to Second World War. These processes were linked to the migrant's life-cycle stage and the “needs” of the individual at the time, not only as the individual perceived these but also as these were perceived by the household he/she was a member of. Thus, the decision to migrate was only rarely taken by the individual [13]. Rather, migration aimed to minimise the risks involved in the sustainability of the household and to maximise its output. In what follows, each of the life-cycle stages outlined above are examined in detail.

The presence of servants in nineteenth-century urban households is evident from a number of sources (Leontidou, 1989, 81; Census returns of Hermoupolis; Census Results of Greece). For the twentieth century there is ample evidence from the in-depth interviews that sending single girls to service was an acceptable way of alleviating household poverty. The usual age at which girls would be sent was approximately 15 years although there are references to much younger ages. While for older girls service meant that they would earn their own dowry rather than extracting it from the parental property, for the very young girls, where only rarely was there a monetary revenue, the purpose would simply be to reduce the number of mouths to be fed in the parental household. In both cases, sending a girl to service meant avoiding the depletion of the parental household resources. Nevertheless, it would occur only when there was excess manpower in the household: “some of my sisters went [to service]. Some others stayed here and did the jobs in the farm. They would produce lime, they would dig, they would thresh” (No 20).

The selection decision seems to have lain with the parents, and especially the mother. It was through informal networks of mothers with Mykoniati in Athens that employment would be found for the girls while they were on the island:

Q. Your oldest daughter you said is B. How old was she when she went as a servant?

A. I think after she finished the primary school…There were some people I knew of and they knew the house…

Q. People you knew of? Were they relatives?

A. No, they were not relatives. But, I had a sister-in-law who was working in that house and she knew that this was a good house… (No 13).

Most of the informants were aware that poor Mykoniati girls were in many cases employed as servants and they saw it as a necessity rather than as a matter of choice. The majority of the informants could name at least one person known to them who had been a servant in Athens. Moreover, there was an abundance of instances where the informant herself or one or more of her sisters had been in service while single. Of 26 informants, nine reported that at least one person within the parental family had been in service. (In the parental family I have included sisters, mother, step-mother and step-sisters).

Being in service had no impact upon their chances of marrying. In all cases where a person was named as having been in service by an interviewee, subsequent questioning revealed that the person did get married. Even though some of the girls met their future husband while in service in Athens, in the rest of the cases the mother would “recall” the girl to Mykonos whenever she thought it was time for the daughter to get married. Since marriage was not necessarily based upon a romantic involvement, marriages could take place very quickly once it was made known that a specific girl was available:

Q. How many years did you stay in Athens [as servant]?

A. I came back 20 years old. Whenever my mother would marry an older sister she would bring back the next one in order to marry her as well. That's what she did. We were ten [sisters]. One after the other [we got married]. She [the mother] kept the order [in marrying us]…

Q. How long did you stay [single] once you came back from Athens?

A. As soon as I came back [I got married]. […] Within a month or so. (No 20)

Thus, service seems to have been not only the sole avenue to migration for a single, working class Mykoniati girl, but also the only way for obtaining regular cash employment, enabling her to accumulate her own dowry and further enhance her opportunities for marriage [15].

Another occupation that could be pursued by single working class women was that of dressmaking. Two of the informants—one referring to herself, the other to her sister—mentioned that they would usually learn the craft by spending approximately a year in Athens as apprentices to an experienced dressmaker. It seems that what determined the employment future of these two girls was the availability of relatives or fellow Mykoniati in Athens who were willing to look after the girls (No 22, No 23).

For single girls of somewhat more affluent Mykoniati families the options were different. One informant indicated that her sister went to Athens to study midwifery while she joined her “to keep her company” and to learn dressmaking. While the former was married and ceased practising midwifery, the informant remained single and she spent approximately 20 years of her adult life in Athens working there: “We would hear, while growing up, we would hear about Athens and we all wanted to go and see Athens and get to know it. That's what it was, really” (No 17). Nevertheless, the informant kept very strong links with Mykonos, where she eventually retired.

This migratory movement to Athens was clearly temporary, especially when it occurred prior to military service in the late teens or early twenties in usual circumstances. The movement took place in the summer, when construction boomed in Athens and there was abundant work for labourers carrying grain in the port of Piraeus. Migrants would return home in the autumn to help their parental family with agricultural work, a crucial time in the Mykonos calendar. Employment in Athens during the winter was scarce: “At that time Athens had no bread [no employment]. The youngsters would go [to Athens] and come back. It was during the summer that we did whatever we could [i.e. earn some money]. Later on when things got better [in Athens] and they could afford living there, then they stayed there” (No 14).

The sole purpose of this temporary migratory pattern was the acquisition of cash, something that was difficult on Mykonos, both for households dependent on farming and for those families residing in the town. The revenues earned by these young migrants—after taking into account their own expenses in Athens—were handed over in most, though not in all, cases to the parental household. The obedience of a son in this respect was taken into account in any “gifts” made by the parents to him either prior to or after his wedding. (These essentially constituted pre-mortem inheritance and were invariably gifts of land (Hionidou, 1995a, 78-9)).

The age of the migrant and the position in the family hierarchy also affected the migratory pattern of a young man. (For a similar situation on the island of Tenos see Dubisch, 1977, 72-3). If, for example, there were younger brothers who could cover the household needs of manpower back on Mykonos, then their stay in Athens in cash employment could be prolonged, thus ensuring a maximisation of the income:

Q. Will you tell me about your brothers? Did they stay on Mykonos? What did they do?

A. They stayed up to the age of 18 years. Then they left to Athens [in the 1930s].

Q. One after the other when they became 18 years…?

A. Of course, they went to work because there was no life here.

Q. And who was looking after the land?

A. The younger ones. And my mother. (No 14)

The same informant indicated that when the children were young, the family owned little land. Only when the children grew up did the family acquire more, presumably through the surplus income earned mainly by the migrant sons. Finally, it seems to have been common for young betrothed men to make a trip to Athens to earn some cash in order to cover part of the wedding expenses (No 7, No 21).

Not all chose to migrate. Still, it is notable that almost all single male Mykoniati had left Mykonos at some point in time in order to work in paid employment [18]. Some spent only a short period in Athens and returned as quickly as possible. The justification offered by informants in such cases was that the migrant did not like Athens (No 11, No 16). Those who stayed on the island almost invariably worked on the parental farm, of which they inherited a large part.

Thus, overall the migratory pattern of single young Mykoniati men was temporary and circulatory [19]. Marriage seems to have played an important role in bringing changes to the occupational status of young men, which in turn had a direct effect on their migratory pattern.

Q. Did [your father] endow all your sisters?

A. Relative to his means, he gave a good dowry to all the girls. Every girl had to get a house with all its humble house-stuff. Some of them sold [the house] and went to Athens, some still have them… when they got married, some of their husbands were working in Athens. They went and set up their households there…

Q. For how many of your sisters did it happen like that? How many of your sisters got married and went to Athens?

A. Five or six [out of ten] (No 20) [Referring to the 1920s and early 1930s]

The decision also depended on whether the couple received gifts and dowry upon marriage, the extent of this gift and/or dowry, or whether they were promised it for a later date. So, in cases where the bride's dowry was a plot of land in Athens, the couple would invariably move there [20]. If, alternatively, the daughter received upon marriage a house on Mykonos, that did not necessarily mean the couple would remain on the island. The house could be sold very easily and the move to Athens would take place if the couple—or in some cases the husband alone—perceived the move as improving their opportu-nities [21]. In cases where a gift was given to the son upon marriage, then he would most probably remain on the island and become a farmer without this excluding the possibility of him migrating temporarily in order to earn cash at a later point. Clearly, the links between inheritance/dowry and migration were very strong.

In the early part of this period, and approximately up to the 1930s, the prevalent pattern was that the family (the mother and the children) remained on Mykonos even if the father had to migrate. This was facilitated by the fact that a house was invariably available to the couple upon marriage, constituting the bride's dowry or part of it. The father either circulated between Mykonos and Athens until he saved enough to purchase some fields, or set up a shop on the island and remained there (No 15 referred to her husband who was circulating for approximately ten years after their wedding; No 3 referred to her father who worked for three years in a quarry in a nearby island). Alternatively, and up to Second World War, the father would circulate between Mykonos and Athens until he retired. (Those working in the 1930s rarely received a pension). A large proportion of the informants' parental families fall in the latter category with the father working in wage employment in the port of Piraeus. Again, in most cases the family owned some farm land on Mykonos (see informant No 14 earlier)—sometimes only a small piece initially—and farming activity would not be interrupted. While the mother and the children were responsible for farming throughout the year, in the most active periods the father returned to Mykonos and performed the necessary tasks (No 13, No 14, No 21 and No 26 referring to their fathers, No 10 referring to both father and husband). These circular migrants would also visit their families during major festivals, but only when they returned for the farming would their stay be prolonged for up to a month as this was unpaid leave from work. Circulation of both single and married Mykoniati men seems to have been very characteristic of the migratory pattern between Mykonos and Athens. This circulatory movement was naturally facilitated by the closeness of Athens to Mykonos and the regular contact of the two places by boat.

Migration of families was also occurring. As time progressed this became more widespread. The most probable reason for the increasing incidence of family migration—in most cases being a newly wed couple—was the substantial improvement of employment opportunities in Athens and/or the lack of any opportunity on Mykonos. At the same time there was an increasing trend of providing dowry in Athens (Stott, 1982, p.206, referring to the pre- and post-Second World War period; also, evident in the interviews).

Strong links between the migrant family in Athens and Mykonos were cultivated from both sides and were maintained throughout life in most cases (For a similar situation on the island of Tenos see Dubisch, 1977). This was achieved first, by annual visits of those residing in Athens to Mykonos for their summer holidays (No 23, No 24) and also reciprocal visits of those residing on Mykonos to Athens for health care or for shopping trips (especially prior to weddings). Second, the links were sustained through the exchange of goods between the families in Athens and Mykonos (No 23, No 21). These goods—in addition to letters and oral messages—would be transferred by private postmen, that is, Mykoniati whose only job was to travel between Athens and Mykonos (No 23, No 13). The large number of these private postmen—21 distinct cases were indicated on birth certificates between the 1930s and the 1950s—suggests the very important extent of the link between Mykonos and the Mykoniati in Athens. (The existence of such private postmen was widespread in insular Greece. For Tenos see Dubisch, 1972, 161-2; also, Kolodny, 1974, 102, 356. Liakos mentions their importance in supplying the Athenian households with servants from the Greek provinces, 1993, 407). Such links proved useful in crisis situations when the breadwinner's ability to provide for his family was impaired (No 23, No 9). In such instances the family would take the decision to return to Mykonos, where “survival” seems to have been easier to secure, as illustrated in the following example:

Q. Were you living in that house [the parental house] until you got married?

A. I got married in the house but then I left and went to Athens… I stayed for some years in Athens and then I came back to Mykonos. My husband became ill and he couldn't work. He was a plasterer. And we came here and he became a private postman. And he was travelling.

Q. Did your husband use to work in Athens before he got married?

A. Yes…

Q. How long was he there? How old was he when he went there?

A. He was there since he was a young boy. As soon as he finished his military service he went to Athens to work… His parents were here [on Mykonos].

Q. Did he use to come back in the summer? When did he come?

A. He used to come in the festivals, because his parents were here. He used to come and go(— phgainoercptane —“meaning going and coming”)

Q. And he met you, you got married and as soon as you got married you left?

A. Yes, we left within a few months [of the wedding] and we went to Athens… They [my parents] gave me [a house] as dowry. And I sold it to my mother [upon migration]. I sold it and went to Athens. And I bought a plot of land in Athens… And I built a house there. I stayed there for seven years and my husband got ill and I left [i.e. the family returned to Mykonos]….

Moreover, some families would return to Mykonos from Athens if and when an opportunity appeared (No 3, No 8). (In the first case the couple returned to Mykonos after spending seven years in Athens in order to open a coffee shop on the island. In the second case the couple returned after 20 years of work in Athens).

The migration situation changed radically in the 1960s once tourism became the prime, and now the sole, industry on the island. New patterns have been established with the island attracting people among whom many are Mykoniati descendants born and/or brought up in Athens/Piraeus. (For a similar situation occurring on another Cycladic island see Kenna, 1993, 75-96).

It would be no exaggeration to say that all young men were involved in migration. Independent of their father’s occupation, migration was an essential start to life when cash would be amassed. The young man’s savings, the bride’s dowry, the potential contribution made upon marriage from his parental family, possible established employment at a destination, along with personal preference, all determined whether the couple would move or stay. Interes-tingly, the mobility of a farmer’s children was identical to that of the rest of the population (No 15, No 2). This reflects the avenues that allowed someone to become a farmer. Clearly the sons of those possessing large plots of land were more likely to become farmers themselves. Still, becoming a farmer could be the end result of years of temporary migration, sailing or circulation, that would enable the accumulation of enough wealth to purchase land and to convert into a farmer either the pro-perty-less wage earner or the owner of a small plot of land [22].

Very extensive was also the migration of individuals originating from better-off families. For this group, migration would in most cases start at an early age for education purposes—either in Hermoupolis or in Athens—and very rarely would they return to the island, except for an occasional visit. With only few exceptions their movement out of the island into high-status employment was permanent. It is the migration of this group that Tsoukalas described. In contrast, among the members of the working class, whether artisans or farmers, employment opportunities in the destination places were restricted to manual jobs ranging from construction to factory work, from mining to laboring [23]. (A similar observation was made for the migrants from Tenos to Athens in the post Second World War period (Dubisc, 1977, 74)). Modifying Tsoukalas’ argument for the working class, I suggest that migration enabled its members to maintain their social class status, thus avoiding proletarianization and an over-dependence on economic fluctuations. Up to the 1920s circulation averted the creation of an extensive proletariat within Greece by the temporary removal of the landless Greeks abroad [24]. At the same time, in rural areas it allowed land ownership to the overwhelming majority of those who eventually stayed on the island.

The patterns of migration showed some continuities as well. The affluent Mykoniati would initiate a migratory move and the working classes would follow. Moreover, the pattern of migration was determined by the life-stage of the individual. As a general rule single men and women, if the household did not necessitate their presence, would migrate on a temporary and/or circular basis in order to accumulate the cash necessary to set up a living either on the island or elsewhere. Decisions to migrate were rarely taken by the indivi-dual. Rather, they were the result of a consensus between the household and individual needs. When married, some men circulated between the island where the family was established and the place where cash employment was avai-lable. Yet, in some other cases a “permanent” move by the couple out of Mykonos would occur. This was a soft version of “permanency”, where strong links were kept with the place of origin and a return to Mykonos would take place in crisis situations or when a particular economic opportunity arose on Mykonos [25]. Gradually, circulation gave way to permanency as employment opportunities became abundant in post-Second World War Athens and social security provision in the form of pensions was becoming reality for the first time (No 9; Sakellaropoulos, 1991, 116).

A very significant finding of this paper is the continual importance of circulation throughout the study period, irrespective of destination. This was the migratory pattern of single males whether they were working in the Danube or Southern Russia in the late nineteenth and early twentieth centuries, or in Athens in the twentieth century; of sailors, whether married or single; of married men working in Athens when their families resided on Mykonos; and of single women working as servants either on Syros or in Athens. Circulation was a risk-minimizing, resource-maximizing strategy for the household and it addressed the problem of absence of social security. Thus, while the family had a residence, some land under cultivation and the immediate support of friends on the island, cash would be earned outside the island. Moreover, the household would invest their earnings on the island mainly by acquiring more land. Part of this land would be given to sons and daughters at marriage, though the elderly couple would retain part of it as a form of “old age security”. Last, but not least, circulation enabled the Mykoniati to maintain a continually updated picture of the opportunities and pitfalls of each destination and thus facilitate or avert a permanent move there.

Circulation was neither confined to Mykoniati nor to a specific time-period. So, the pattern of migration of Greeks to Constantinople in early nineteenth century was a circular one for the migrant while his family resided in the village of origin (Liakos, 1988, 45 referring to a detailed study of patients in the Greek hospital in 1833-39). In mid-nineteenth century Piraeus, the workers were mainly young single males, involved in seasonal/temporary migration “coming and going” (Mpafounes, 1984, 563). In the highlands of Western Macedonia circulation was a well-esta-blished practice throughout the nineteenth century while by the turn of the twentieth century the peasants of Wes-tern Macedonia got involved into a circular migration to the USA (Gounaris, 1989, 134, 149). Circular migration of males —single or married— was also “traditional” in the highland Zagori villages in Epirus, at least since the late eighteenth century and throughout the nineteenth century (Campbell, 1979, 12-3). In the 1905 manuscript census of such a village, 310 men out of a total of 418 were recorded as “absent”, suggesting that circular migration there survived into the twentieth century (Caftanzoglou, 1997, 412-3). In late nineteenth and early twentieth century the labour shortage in the port of Patras was filled by seasonal migrants not only from the surrounding areas but also from the Ionian Islands and the mountainous interior of Peloponnese (Frangakis-Syrett and Wagstaff, 1996, 129, 133). For extensive seasonal migration to occur, most probably, many individuals participated for more than one season. Seasonal migration from the Ionian Islands to the coastal areas of Epirus in the late nineteenth century was also occurring (Psuchogios 1985, 27). Almost all married Greek migrants to USA left their families in Greece while 61 per cent of those who migrated in 1908-1930 returned home (Tsoukalas, 1987, 150-1). A circulatory migration pattern of married and single men to USA was also described to the author in interviews of inhabitants of the Aegean island of Chios. Similarly, the author of the autobiographical novel “Eleni” was left in Greece along with his siblings and his mother while his father was working in the USA prior to Second World War (Gage, 1989, 443). As Tsoukalas puts it “the Greek migrant [to the USA], just like his grand-father and his father migrated to Romania or Russia, leaves to return, always remaining free to migrate, if the necessity arises” (Tsoukalas, 1987, 150). Finally, Arnold Costa mentions the existence of an “elaborate circular seasonal migration pattern” after 1960 on one of the Ionian Islands, Cephalonia (Arnold Costa, 1988, 180). Currently, circulation is still in full operation though taking different forms. For example, workers of the tourist industry originating either from Athens or from villages and towns moving to the resorts for the tourist season or the large numbers of islanders who own a residence in Athens and many of who usually spend the winter months there. So, circulation seems to be “a time-honored and enduring mode of beha-vior” persisting over-time and over a radically changed economic background (Chapman and Prothero, 1977, 5).

Thus, while the evidence remains scant, it is growing pointing to the existence and continuity of circulation. I suggest that circulation was the most important pattern of migration among the members of the Greek working class both in the nineteenth and in the twentieth century.

Circulation has been extensively stu-died in recent years for less developed societies ranging from Africa to Latin America and Indonesia, referring mostly to the second half of the twentieth century (Elkan, 1967; Arizpe, 1981; Hugo, 1982; Prothero and Chapman, 1985; Forbes, 1984). Much less attention has been paid to circulation in historical studies, the primary reason being the difficulty of establishing its existence with the available data (bright exceptions being Moch’s work on Wes-tern Europe and that of Borges on Southern Portugal). To what extent circulation was prevalent in the rest of Southern Europe and the Balkans historically remains to be explored.