The Life and Death of Residential Room Types: A Study of Swedish Building Plans, 1750–2010

While the study of building types is a well-known and relatively active research field, the topic of room types is less explored. This article takes a broad approach to spatial categorization, enabling the examina - tion of different types of spaces over longer periods. How do different room types evolve and die? How do the different residential room types relate to each other? Do they act alone or do they follow each other over time? The article looks at the particular evolution and development of Swedish residential room types and is based on the study of plans of 2,340 Swedish buildings from about 1750 to 2010. Six themes emerged from this study: thresholds of birth and extinction, abruptive change, the relation between absent and present room types, contagious types, different temporal scales and the stabilization of prototypical sets.


Introduction
Spatial entities can be classified into different types of rooms.These types are often used in building programmes and briefs (Markus and Cameron 2002), and to set the plan of buildings, buildings that subsequently are aggregated into urban areas, and urban areas into cities.Room types here play an important part in how we behave in everyday life (for example, justifying certain restrictions, such as ' quiet, this is a reading room'), and they take part in the transformation of objects and cultures of different scales.Like the classification of building types, the naming and designing of room types is a matter of territorialising specific kinds of spaces (Kärrholm 2013), and as such, types of rooms participate in the controlling and ordering of movements and behaviour (Sack 1986).However, whereas the study of building types is a well-known and active research field (Markus 1993;Forty 2000;Scheer 2010;Guggenheim and Söderström 2009;Steadman 2014;Karlsmo and Löfgren 2016), the topic of room types has received less scrutiny.Research on residential room types is so far a quite fragmented field, encompassing everything from general and specific design guidelines (Neufert 1936) to more descriptive and historical writings on room types (Barley 1963;Muthesius 1979;Gejvall 1988).It also includes research on the relationships of room types and their spatial distribution within dwellings (Hanson 1999;Nylander 2013).In general, literature on residential room types has often focused on a specific historical, typological and geographical setting, such as large country houses in Sweden (Selling 1937), England (Girouard 1978) or Ireland (MacCarthy 2016); bourgeois apartments in Stockholm during the 19th century (Gejvall 1988); Victorian homes (Girouard 1979;Flanders 2004); or the room types of specific rural contexts (Erixon 1947;Barley 1963;Hansson 1999).
I look at the evolution and development of residential room types and how they relate to each other.The article can thus be seen as an initial investigation of some general themes of room type transformation: How do different room types evolve and die?How do different residential room types interrelate?Do they act alone, or do they follow each other over time?Sweden provides an interesting case study, since the country underwent an unusual, quick, thorough and dramatic modernization (and urbanization) during the 20th century, and so trends in transformation can be easily identified.However, this transformation is set within the broader historical context of the modern era of architecture, starting (as suggested, for example, by Collins 1965) around 1750.The article is thus based on a study of 2,340 building plans of Swedish buildings from about 1750 to 2010.It explores different themes that emerge about the transformation of room types and ends with six ways that room types intermingle and come and go.

Spatial Types, Room Types and Territorial Sorts
In this article I discuss type as a spatial category that matters in everyday use.I do thus not follow the morphological conception of type, sometimes called form type, that can be found, for example, in the famous studies of Jean-Nicolas-Louis Durand (2000Durand ( [1802-5]-5]) and Saverio Muratori (1959).Since form and use must be studied together, it is better not to employ the more common notion of use type (Scheer 2010: 10;Steadman 2014: 354).Rather, I follow Steadman's more general definition of building type as ' a classificatory unit by which similar buildings can be grouped and enumerated ' (2014: 353).To this definition I would add a more pragmatic perspective: a type is also always a kind of actor, something that has an effect (Latour 2005) in an everyday life situation.A certain type of space, such as a bathing place, might come in a variety of different forms and host a series of different functions (Carl 2011).However, it is because someone recognizes it as a bathing place and uses it accordingly that it makes a difference in our everyday life.Both social (who can bathe and how?) and material aspects (what kind of bathing does this specific place afford?) have their role to play, and these roles are interdependent.Issues of form have often been distinguished from issues of activity or use, so that the difference between single-family detached houses and row houses has do with form, whereas that between student housing and elderly housing is about use.One problem with putting a focus on either use or form is that it tends to omit buildings without clear purposes or with irregular forms (Karlsmo and Löfgren 2016: 12). 1 Also, when we look at transformations, there is always (as we shall see) a change both in form and use.The room type of the Swedish kitchen, for example, has changed in both form and use over the centuries.What it can be used for and by whom, as well as its form, location and integration in the house has changed many times, yet it has retained its identity as a kitchen.Types can also be described as 'territorial sorts', since they territorialize a certain object or space with a certain meaning/intensity (Brighenti 2010;Kärrholm 2013).Types are not innocent, but are soaked in power.They are in fact a way of turning a certain space into a socio-material actor with a certain effect, i.e., into a territory.One could also describe them as 'sorts' to make clear that they are not defined by a standard set of entities (like prototypes might be), but must be seen as a more fluid assemblage where no entity is in itself obligatory.Instead, different entities can come and go over time as long as they share some kind of family resemblance (Law and Mol 1994).
So, how do we approach the question of typological transformation?In 1825, Quatremère de Quincy defined an idealistic concept of type: ' an object after which each [artist] can conceive works of art that have no resemblance' (Quatremère de Quincy in Steadman 2014: 353) but instead have an elementary principle in common.Ever since then, type has often been used from a normative and prototypical perspective (Steadman 2014: 354).In Aldo Rossi's L'architettura della citta (1966), for example, identifying recurrent types of buildings was a way to justify architectural form, because recurrent types ensured a certain meaning, producing a historical continuity within the city (Forty 2000: 304-11).Urban morphologists as well as urban and architectural historians have also used a more empirical notion of type, but nevertheless remarked on its form as a mental image.For example, Caniggia and Maffei argue that types can be defined a posteriori, but they also claim that their existence actually depends on the fact that 'it [the building type] exists in the builder's mind before producing a house' (Caniggia and Maffei 2001: 53; see also Kropf 2001).
Types, or territorial sorts, are abstractions that enable us to think and do, but they should not be seen as ready-made solutions or an ordered list of rules.Territorial sorts are often too dependent on vague associations, atmospheres and affects to be formalized as some sort of mental model or a rule of thumb.As crime fiction has taught us, a gunshot in the dark might change one type of place into another, an idyllic village street into ' a dangerous and scary place' or even into ' a crime scene' in the blink of an eye.To name or categorize something as a specific type or sort of space is thus a quite basic phenomenon, and should not be confused with the much more specific case of typification that we see in modern and industrialized housing (where room types might be sorted into taxonomies, and where each type might be defined more formally through different kinds of regulations).Formalization is an exception rather than a rule when it comes to the effect of types.I would therefore like to suggest a much more fluid view on spatial types.They might of course hold a certain stability, but arriving at a definition is always a struggle, because all types are always on the way towards something new (Kärrholm 2016).In short, when it comes to types, there is always continuity as well as continuous change in both form and use (Koch 2014).
Typological transformations can be investigated through comprehensive historical and ethnographic studies (e.g.Paulsson 1950).Although I advocate studies of this kind, I also consider it necessary to address the question on a more abstract and general level.One way to do this is through a biological analogy (Kropf 2001;Steadman 2008Steadman , 2014)).The analogy between typological transformation and Darwinian evolution was often taken to mean that types improved and became more complex and advanced (with a 'better fit') over time.This view was especially popular during the late 19th century (Karlsmo & Löfgren 2016;Werne 1997), but can actually also be found in more recent building type theory (Scheer 2010: 27).The theory of evolution does not, however, state that there is a pre-determined hierarchy of types (in terms of value), nor that types always develop from basic to more complex ones -both directions are always possible.The development of room types or building types seldom occurs through random variations (Steadman 2014: 3).Neither are such types provided by a certain environment or context in any deterministic sense.
Nevertheless, a metaphor of spatial species, or more specifically, room species, is a first step towards a more animated and ecological discussion of types.Darwin himself was well aware of the problem of defining a species (Darwin 2011: 44-50), and species are always, as Brighenti suggests, 'both territories and movement' (Brighenti 2014: 11).Species are constantly moving figures, and categorization is always a temporary abstraction.Evolution theory has shown us that we are machines of difference (cf.Deleuze 1994).Life is continually producing differences, and if selection (random or not) seems to temporarily stabilize a species, enabling an abstraction, the forces of deterritorialization are always working in its very midst.This process can, for example, be discussed in relation to the visitor centre type of building (Kärrholm 2016).The establishment of different kinds of information and welcoming spaces relating to tourist attractions led to the development of the visitor centre as a new building type in Sweden during the early 1990s.However, only a decade or so later it seems as if different subtypes were formed, bearing the seeds of several new species of space.Theoretically, every individual difference always has the potential to be the start of a new species.Room types, involved and entangled as they are in our everyday lives, are no exception.

The Study of Building Plans
A large base of empirical material, preferably covering a long period, is necessary to study the evolution of room types.The empirical material for this article is based on the study of the plans of 2,340 buildings made in Sweden or drawn by Swedish architects from around 1750 until 2010.Of these, 816 buildings are residential types and 1,524 represent building types primarily built for other purposes (hospitals, schools, governmental buildings, railway stations, museums, etc.).The residential types were more important for this article, but residential room types in Sweden were also found in public buildings, factories, office buildings, etc., until at least the 1960s. 2  In the study, all of the different room types in each building plan were noted and sorted chronologically according to building type.The database consists of a total of approximately 40,000 to 45,000 rooms.The building plans were collected through an inventory of all issues of the Swedish magazine Arkitektur (the Swedish architectural review and the largest Nordic magazine about architecture), which began in 1901. 3To better account for the years before 1901, a number of reference works on important building types and architects was also used (see reference list).Information on room types between 1750 and 1900 is also taken from the well-documented architectural history of residential buildings in Sweden (e.g., Selling 1937;Lundberg 1942;Erixon 1947;Gejvall 1988;Nylander 2013).
The homes of the middle and upper classes are well represented because these sources focus on buildings drawn by architects and published in architectural journals.Some room types might therefore also be omitted altogether, such as antiquated but enduring rural room types and room types common in poorer housing, like the spisrum (literally, stove room), found in and around Stockholm in the second half of the 19th century.In single room apartments, the spisrum was a combined living room, bedroom and kitchen.Although this type was quite common, it does not appear on any of the plans I studied.
The work also relies on terms found in the Swedish national dictionary, Svensk Akademisk Ordbok (SAOB), an ongoing project that began in 1893, as well as in the shorter but complete and more updated Svensk Ordbok (SO). 4 While naming is of course an important part in the spread of spatial species (Steadman 2014: 360), it should be remembered that these dictionaries are based on texts and not on plans, and actually account for words used for room types whether built as such or not.

Themes in the Life and Death of Residential Room Types
Rather than presented in a strictly chronological way, the findings are explored as a series of different themes.Tables 1 and 2 provide background for the discussion, since they present the approximate lifespans and dates of origin of a number of room types.Room types often feature in everyday language long after they have ceased appearing in plans (during my own childhood in the 1970s and 80s, it was thus not unusual for older people to refer to a room in their apartment or villa as a maiden chamber, even though the room had not been used as such since the 1940s or '50s).A room type might also appear both as a concept and as an actual place before it receives a specific and deliberate design by an architect (cf.Steadman 2014: 368).

Thresholds of Birth and Extinction
In the plans for the large Swedish houses of the 17th and early 18th centuries, rooms were often classified according to size, and not always according to function (Gejvall 1988: 171;cf. Rybczynski 1986: 42).Houses included kammare (small chambers) and kabinett (cabinets), and salar (larger rooms) and salonger (salons), and the interior arrangement of these rooms was often important.Larger bedrooms could, for example, have both antechambers and smaller back rooms to which inhabitants could withdraw (Figure 1).The sequence of movement and the notion of hosting guests was thus an important matter in these plans (cf.Baeckström, 1917: 46).
The chambers and cabinets, however, soon grew more specific -porcelain chambers, milk chambers, writing chambers, guest chambers, etc. -while the more general types of chambers and cabinets gradually disappeared.The early Swedish texts on architecture, such as those by Johan Eberhard Carlberg (1740), Carl Wijnblad (1755-56) andCarl Stål (1834), were all greatly influenced by French architecture, and Swedish residential plans often followed the French style of organizing prominent houses, with an enfilade of predominantly large rooms towards the front of the building, combined with two sets of rooms (antechamber, bedroom, cabinet and wardrobe), one for the husband and one for the wife (Figure 2).These were principles that French architects such as Augustin-Charles d'Aviler, Charles-Étienne Briseux and Jacques-François Blondel introduced during the 17th and 18th centuries.An interesting Swedish example can, for example, be found in Baron F. Löwen's house in Stockholm, from the 1740s (Gejvall 1988: 105f).
The earlier Swedish houses and apartments, influenced by French architecture, could thus be seen as divided into two parts (Gejvall 1988: 255).One part contained social rooms and living quarters, with a large room -the sal -in the middle, and with one series of rooms on the husband's side and one on the wife's side.These two different suites of living quarters often included a förmak (antechamber), a bedroom and a cabinet.The other part of the house contained the kitchen area with servants' rooms and possibly children's chambers.
The great room, or sal (quite similar in use and style to the French salle; Rybczynski 1986: 38) was an important Swedish room type that acted as a living room, a dining room, and a room for work as well as for parties and for hosting guests.It was also an important room for movement, as many apartments actually required all visitors and family members to pass through the sal to reach the other rooms of the apartment (Gejvall 1988: 188-200).The room type tambur, a kind of lobby or antechamber, was introduced in Stockholm apartments at the end of the 18th century, and came to act as a kind of small entrance room to the sal.It was often used to hang clothes and store wood for the fireplaces.Later it became an important connector to other rooms as well, and as such was well integrated in the spaces of the home.The idea of an antechamber to main rooms was a French idea, but the role that the tambur soon took on was probably more influenced by the way that the Vorzimmer was used in German apartments during the 19th century (see Gejvall 1988: 175-86 for a more thorough discussion on the tambur).
The importance of the tambur increased, and it become larger and lighter, growing into a kind of lobby during the second half of the 19th century (Figure 3).The tambur also came to be physically connected to such increasingly popular types as serving rooms and linen rooms, as well as to corridors and passages, and as such it became key to enabling movement through the dwelling without passage through the more presentable, formal rooms (Gejvall 1988: 107f).The role of the tambur thus changed.It, together with the sal, takes on a less presentable and more functional role, becoming a passage, while the sal often is reduced to the function of dining room (mat-sal).Instead of the sal, which became less central over time, the main formal rooms now became the salong and the förmak.The spatial organization of the late 19th-century Swedish apartment, inspired by the continental apartment in countries such as France, and interestingly also Germany, can, according to Gejvall (1988: 255), be described as divided into four areas (Figure 3):   1942 247).The distance between these rooms thus tended to be extended during the second half of the 19th century (Gejvall 1988: 199).
From the 1890s onwards, the English influence grew stronger, and the hall started to become a more important and more dominant room (Lundberg 1942: 257;Gejvall 1988: 168ff;Paulsson 1950: 120).The hall plan became popular in the new middle and upper-class villas of the  late 19th century, where rooms were arranged around the hall, rather than in line.Soon, the tambur was exchanged for a hall and a cloakroom.The cloakroom allowed the hall to be free from clothes and other paraphernalia that often had cluttered the tambur (Figure 4).The hall, which was more of a room to dwell in than the tambur, also opened up for the spatial connection to even more rooms (of different types), such as the vardagsrum (approximately equivalent to living room, but literally meaning ' every-day room'). 5At this time, there was also a critique of the salong/förmak (Gejvall 1988: 224) and the idea that hosting visitors seemed to be more important than the living conditions and comforts of the inhabitants.Perhaps this was one of the reasons why the centrally integrated hall, the cloakroom and the vardagsrum became popular, whereas the salong and the förmak slowly disappeared (Figure 5).In Sweden, the number of residential room types also seems to peak around 1910 with Isak Gustaf Clason's houses.Clason designed a series of large houses during the decades around the year 1900 (the last of the large Swedish estates of the 19th-century tradition), still following the Victorian tradition and its plenitude of room types (Edestrand and Lundberg 1968: 48-62).After the 1920s, the number of room types seems to decrease, and during the 1950s several of the former important types were gone altogether (Figure 6).In non-residential building types, the proliferation of more and more specific room types went on longer, until the 1950s.One example is the hospital Sahlgrenska, in Gothenburg, which had about 170 different room types.Perhaps it was only with structuralism, and the call for flexibility during the 1960s (Forty 2000), that the decrease in room types became a general trend.
As the functional differentiation of the home increased during the 19th century, so did different kinds of power asymmetries.Moving the living quarters of servants, as well as kitchens, bathrooms, etc., into the home also played a part in the co-production of distances and asymmetries within the home itself.Social distinctions were important and thus came to take an architectural form (Rybczynski 1986: 49), where servants' rooms were named according to the title or category of the servant.The builtin asymmetries at the room type level seem to decrease as servants moved out and as the number of room types declined in the mid-20th century.
In summary, residential room types proliferated, first through the French idea of different suites of rooms, then on through the continental apartment plan and the English hall model; all three different models made it possible for the number of room types to increase.The first wave of extinction started slowly during the 1910s and '20s, with the fall of the representative room types.A second wave of extinction was around the Second World War.The linen rooms, brushing room, maid's chambers, and gentleman's rooms 6 all disappeared during the 1940s; the tambur and the serving room declined quickly at the same time but endured for a few more decades before disappearing altogether.The tambur, which perhaps was the first catalyst for this rich tradition of different room types, was thus the last to go.Following this history, we can observe that spatial species do not always come and go one at a time.Instead, it seems that there are quite often thresholds where series of new spaces evolve or die.This has been made clear in studies of building types that show, for example, that industrial society came with a series of new building types concerned with production (Markus 1993), but these thresholds seem less studied when it comes to room types.This change of the mid-20thcentury residential spaces was perhaps most of all about fewer people and more things (Westerberg and Eriksson 1998: 268); as a welfare society developed and the home became a place for the family to host things rather than people, room types connected to an older kind of society started to decline.

Abruptive Room Types
Not all room types come and go in groups.Indeed, a type might also arise from a sudden, abrupt or disruptive invention, a ' chronic' moment (Brighenti and Kärrholm 2019) of upheaval and change.Here, I will just briefly mention two examples of room types where this more  subversive aspect (breaking with former tradition), is stronger: namely the divan room and the allrum (family room, or literally, ' everything room').
The divan as a piece of Ottoman furniture was popular in Europe from the last decades of the 18th century and into the first half of the 19th century.Sweden had good diplomatic connections with Turkey during the 18th century (Avcioglu 2011: 255ff), and King Gustav III made several divan rooms in his palace at Haga (for example, in the pavilion drawn by Olof Tempelman in 1787).The divan thus migrated into Swedish architectural culture and got its very own room.In fact, in Swedish, divan did not refer to the furniture at that time; in a Swedish dictionary from 1853, a divan is defined as a room with low sofas along the wall (Gejvall 1988: 241f).Most often the divan room was a bit smaller than the salon and could be found in large private apartments or houses, like at Stora Bjurum, where a divan room was introduced in 1869 (Figure 7).In at least one case, however, the divan room was also used in a public building.In Helgo Zettervall's first plans for Sweden's Parliament House and Central Bank building in 1884, he placed a large divan room very centrally and in direct relation to the foyer (Bodin 2017: 815).This room disappeared in later plans for the building, and to my knowledge it has not appeared on any Swedish plans since then.As a room type, the divan room has no clear (Swedish) predecessor, and it disappeared almost as quickly as it appeared.The second example of abrupt inventions is the allrum.Although the vardagsrum had replaced the salon during the early part of the 20th century, it appears as if this room also became used for more formal events and the reception of guests, and thus the struggle of architects and politicians to establish a room for everyday use continued.The importance of a space used for display and hosting guests, for example, often meant that sleeping facilities were less prioritized, and this was perceived as a societal problem (cf.Paulsson 1950: 120;Perers et al. 2013).One important attempt to change this was the allrum (literally, ' everything room'), a new room type that was introduced in the housing competition of Baronbackarna in Örebro 1950Baronbackarna in Örebro (built 1953-57)-57).The housing competition was not aimed at inventing a new room type, but the winning proposal, announced in 1951 (and designed by P.A. Ekholm and S. White), suggested that the traditional living room could act as a kind of study room or studio (arbetsrum).On later drawings this room was referred to as an allrum.An experimental apartment with the very first allrum was built during the summer of 1952, and furnished with a kitchen sofa, a dinner table, a bookshelf and a desk with a sewing machine (Krantz 1987: 96ff; see also Mack 2017: 228ff).The new room type received a lot of attention among architects (Figure 8), but studies in 1956 showed that the introduction of the allrum initially failed.Rather than being used as the family room, as intended, it was used as a kind of salon or finrum (literally, 'nice room'), and it was only with the introduction of TV sets during the early 1960s that the allrum became a more everyday kind of space (Krantz 1987: 103).Its popularity increased from the 1970s onwards, finally making the idea of a salon or a finrum (even if enacted within spaces tagged as living rooms) obsolete.

Absent Friends and the Ever-Changing Boundaries of the Dwelling
In his book Objects of Desire (1986), Adrian Forty has argued that the modern Western home is generally a product of the Industrial Revolution.Through the development of specific work places, coupled with the regulation of work and work hours, the home soon became an important haven of privacy, comfort and leisure: an antipode to work (Forty 1986: 99).During the 19th and 20th centuries, the boundaries of the home were also changing.Rooms and functions that formerly were located outside the dwelling itself, like places for food, bathing, laundry, latrines, etc., now moved inside the dwelling (Gejvall 1988: 101).From a situation where rooms could be rented out even without kitchen facilities (so-called bachelor flats), the standard and comfort of living thus slowly began to increase.New technical infrastructures, such as water pipes, were also introduced in several Swedish cities during the 1860s (Paulsson 1950: 185), and the elevator, which afforded an easier movement of goods and people in and out of the home, was introduced during the 1880s.
New technology developed between around 1880 and 1920 made it possible to achieve a degree of domestic comfort without servants and the accompanying spatial separation between room types.Some room types were no longer needed, or their functions could be integrated into other rooms.As we have seen, the number of 'indoor' room types seems to decline from about the 1920s and onwards, and the decrease in secondary spaces outside the home proper actually also continues slowly throughout the 20th century.During the 1930s and '40s, storage spaces in cellars decreased (Gejvall 1998: 212), as refrigerators and freezers became more common and fresh food could be bought all year round.Another function that moved out of the home was child care.The building of Swedish day-care centres in the 1940s and '50s meant that child care in the home or in other people's homes decreased.Bringing laundry facilities closer to the home was another important issue during this century.Collective laundry rooms with washing machines were introduced from the mid-1920s by HSB (a Swedish cooperative association for housing); a national suggestion for municipal regulation requiring laundry rooms was made in 1948 (Figure 9), and in 1965, 90% of the Swedish population had access to a laundry room (Björkman 1985: 88;Lund 2009).However, in the 1990s a series of neighbourhood disputes centred around these places, and as technical developments were made, washing machines moved into apartments (into kitchens and bathrooms).
Parallel to the privatization and movement of activities and goods into the home, the possibility to actually harbour all of these objects and functions had grown increasingly problematic.Although the rising problem of storage space was noted already in an investigation in 1980 (Konsumentverket 1980), the decline of community spaces in the large residential housing areas continued during the 1980s and 1990s, and storage spaces today have to an extent been commercialized and outsourced, as new companies specializing in storage facilities have developed, often in former industrial areas (Brembeck 2019).
Dwellings changed as new room types moved in and others moved out.This can partly be explained by new technology (like water toilets, freezers, etc.), as well as an ongoing urbanization in combination with higher living standards and increasing consumption, which meant that storage spaces were externalised.Activities that were once performed in the home are now executed in other neighbourhoods, cities, regions or even countries (sites of production that were formerly in the home might, for example, have moved to the other side of the world).The home and its room types thus co-evolve with the environment in which they are located.The absence or presence of the different activities of the home seems, for example, to be related to longer trends such as technological development, transformations from a rural to an urban society and globalization.

Contagious Room Types
Related to yet different from the notion of absence or presence is the exchange of room types between different building types.Until the 1960s, it was not uncommon for the homes of rectors, cleaners, teachers, drivers or janitors to be integrated into public buildings such as schools, museums, public baths, etc. (Figure 10).Residential functions do, however, disappear from public buildings with the growing tendency to see the home as a protected and individualized unit separate from work, a place exclusively reserved for leisure and a nuclear family life.The sharing and exchange of room types between residential and non-residential building types nevertheless continued.For example, the vilrum (resting room), which was established as a Swedish room type at workplaces during 1940s, appeared in residential buildings during the 1970s.Similarly, as larger structures and building complexes became more common in the modernistic large-scale plans of the 1960s and '70s, more traditionally urban categories such as 'square', ' area' and 'street' began appearing indoors.This is connected to the state-subsi- dised large-scale housing projects in Sweden that began in the mid-1960s, when housing blocks and houses took on a whole new scale (Andersson 1976;Hall & Vidén 2005).
Influenced by Team X, large corridors were for example called 'interior streets' in Bengt Edman's Sparta student housing in Lund in 1971.The suffix -yta (area) also became more common, such as arbetsyta (work area) or lekyta (play area) (Figure 11).Indoor squares may not be found in residential buildings, but they began appearing quite often in schools, churches, commercial buildings and offices.There are also examples of how residential room types move into non-residential building types; for exam-  ple, the allrum soon become popular in schools, and the hobbyrum appeared in churches and parish halls from the 1960s and up until at least the 1980s.

Trends of Different Temporal Scales
The evolution of room types proceeds at no predictable pace; some develop slowly and over a long time, whereas others might bloom during a short period of time and then fade away.Old and new room types can thus always be found side by side.One example of a short-lived trend was to refer to room types in a diminutive form, by exchanging the suffix -rum for -vrå (nook).Due to the density of the population and a housing shortage, there was in interest in the concept of small dwellings during the late 1920s and '30s (Björkman 1985: 97), when small room types also developed, also with the suffix -vrå.The kokvrå (cooking nook, or kitchenette) makes its first appearance in 1927 as the most long-lived and important of these types.
Other examples include the frukostvrå (breakfast nook), introduced in 1920, sittvrå (sitting nook) 1925, bokvrå (book nook) 1927 and arbetsvrå (study nook) in the 1940s (Figure 12).The suffix did not take off to the extent first suggested, but it was not a total failure either.Some of the nook types still exist, and new versions were also tried later (like the short-lived tonårsvrå (teenager nook) in public libraries during the 1960s).
Another trend that peaked between the 1950s and the 1980s was room types related to leisure, consumption and free time.For example, although it already existed as a room type, the sauna suddenly became more popular during the '50s (Figure 6).Rooms for hobbies, ping-pong and weaving also appeared during these post-war decades, as well as storage spaces specifically for sports equipment like skis and sleighs.Some rooms might thus only live or thrive for a couple of decades, but there are also room types that disappear even quicker.The 'battery room for the door bell' (on a plan from 1870) is one such very short-lived room; another one is the bodega (Figure 13).These shortlived rooms are quite often (but not always, as the case of the bodega shows) related to new technology, and stabilized through what can be called network stabilization (Law 2002); they depend on a series of obligatory actors, such as laws, certain technical infrastructures, etc.One example of this is the telephone room.
The long-lived room type, on the other hand, evolves slowly and mutates, like the garderob (wardrobe), which over time changed almost beyond recognition (Figure 14).In the plans of Jean Eric Rehn's Lambohov of 1762-66, the garderob is a quite spacious through-way room with a window and a tiled stove.Throughout the 19th century, however, the Swedish garderob was a small, often dark, walk-in closet.It was quite popular, and there were often more wardrobes in Swedish flats than in corresponding flats on the continent (Gejvall 1986: 240).Today, walk-in closets are more often referred to as klädkammare (clothes chamber), whereas garderob more often refers to a standardized, built-in piece of furniture for clothes (cf.Rybczynski 1986). 7 Other very long-lived room types manage to resist evolution and to some extent remain the same over centuries.A case in point is the kitchen.Kitchens may have changed from closed to integrated, small to large, but they are still recognizable as kitchens.The kitchen relies on a more fluid stabilization (Law and Mol 1994); i.e., actors may come and go, new ones may be added, others may disappear.Fluid stabilization does not rely on a specific set of actors, but more on a family resemblance between actors.The associations between actors might change -as long as changes are not too abrupt, new actors might be welcomed or released from 'the family'.One way of illustrating the fluidity and versatility of the kitchen as a category is through its many variants.In the plans studied, from 1750 to 2010, at least 35 different kinds of kitchen types appear, including the quite common grovkök (literally, 'rough kitchen', which is the Swedish term for a scullery), sandwich kitchens, milk kitchens, children's play kitchens (Figure 15), training kitchens, tea kitchens, paint kitchens (färgkök, found in theatres), and barium kitchens (found in children's hospitals). 8

Prototypical Sets
During the period 1750 to 2010, Sweden developed from a rural to an urban society less focused on production and more on leisure and comfort.In cities, residential room types such as the dining room, salon and gentleman's room, thrived and evolved during the 19th century.However, many rural room types disappeared during the same time (cf.Erixon 1947).The first Swedish norms and recommendations around housing, and its minimal requirements, were published in 1921, and a more proper Swedish science around housing started with the foundation of the research institute on housing (Hemmets forskningsinstitut) in 1944.These efforts eventually led to better living conditions, but also to standardization, and to fewer and more uniform room types.In fact, the number of residential room types appears to have decreased over the centuries, reaching its minimum with a small set of general (and on plans often nameless), standardized room types in the 1960s and 70s.
The stabilization of a series of aligned room types (first by research and then by state recommendations and legislation), is related to the home as a stabilized type.The home grows more and more stabilized as a place of comfort and retreat, but also of functional efficiency.The increasing research on residential room types during the 20th century -how to design kitchens that were easy to work in (Thiberg 1968) or bathrooms that were easy to clean (Linn 1985), etc. -together with the decrease in the number of room types, also paved the way for the standardization of a set of obligatory room types (bedroom, kitchen, bathroom, living room).The home became a type produced 'from a standard "kit of parts"' (Steadman 2014: 358).
The general decline in the number of room types also means that some rooms had to take in activities which had previously been performed in a set of other rooms; the bathroom, for example, now includes activities which formerly might have been done in boudoirs, laundry rooms, nurseries and dressing rooms (Rybczynski 1986: 223).The bedchamber, the children's chamber and the guest room were all transformed into the more standardized 'bedroom'.From the 1960s and 1970s, the names of residential room types also began to disappear from plans in the journal Arkitektur.Since these room types were now rather few and tended to be the same from house to house, year after year, there was no need to mark them out.In the early 1990s, however, this changed.The Swedish building standard for room sizes was abandoned, and the building code no longer dictated a detailed prescription for the layout of housing plans, which in turn opened the way for new room types, as well as a new fluidity when it came to the existing ones.The names of room types on residential plans become somewhat more common again, and new types, such as the relaxation room and the spa, have appeared, a trend that seems to continue during the 2000s.

Conclusions
The aim of this article has been to explore an evolutionary approach to room types, arguing that the historical development of room types should not be studied on the basis of single entities, but must be understood in relation to other room types and their ecology -in the house and elsewhere, and at other times.From the study of Swedish residential plans, I have derived and discussed six themes that are most probably not unique to Sweden (even though they sometimes took a particular form in Sweden), but are similar to trends in other European countries.After all, Sweden was entangled with the development of residential cultures in other countries -like France during the 18th century (with an emphasis on rooms such as the salon), Germany (with the influence of the Herrenzimmer) and later Britain (with the hall plan) during the 19th century.Room types tend to come and go in packs, and they seem to be subject to thresholds of evolution and extinction.Room types can, however, also be more abruptive.Although room types often are associated with each other during longer times, single room types can also come and go suddenly (like the divan room or the allrum).Room types are also always dependent on 'absent friends'.The room types associated with a certain place, such as the home, are produced in relation to the environment of this place.There is thus a constant exchange of room types as the home and its environment are co-produced over time.Furthermore, some room type trends are contagious.Room types do not stick to specific building types but can spread between different contexts and be reused or reinvented across different building types or urban types (for example, between homes, schools and churches, between cities and buildings, etc.).Series of rooms also follow different temporal scales, that is, old and young room types often live side by side.Room types that are stabilized through a series of obligatory actors seem to be more short-lived than the more fluid territorial sorts.Finally, room types can stabilize into prototypical sets, the obligatory room types that tend to make up a certain place or building type.
It is fruitful to discuss room types as a kind of spatial species.Rather than separating them into use types or form types, we address them as important and transformative actors in architectural and societal production.Individual differences always have the potential to be the start of a new species and if we are interested in studying this change -types on the move (cf.Latour and Yaneva 2008) -we cannot reduce the notion of type to a certain category, but need to follow all the different ways in which it makes a difference.When looking at how types transform, the question is thus not whether they are defined by form or use, but how they have an effect on ongoing life, how they keep or change their identity and how they evolve, decline or even die.

Notes
1 On a more general note, a focus on types might also risk a focus on spaces as objects, obscuring situations and practices (see, for example, Carl 2011 on the relation between 'type' and the 'typical'). 2 At that time, there was no longer any need for someone to keep an eye on the building during evenings and weekends (keeping the fire going to heat the building, etc.).Housing became more affordable and readily available during the 1960s, and the idea of 'the home' was also changing rapidly during the post-war decades.Home became a place of leisure, and more firmly separated from the place of work (see, for example, Forty 1986). 3The journal Arkitektur was published in Stockholm by Arkitektur förlag between 1901 and 2010, and appeared under the following titles: Arkitektur och dekorativ konst, from 1901to 1908;Arkitektur, from 1909Arkitektur, from to 1921;;Byggmästaren, from 1922Byggmästaren, from to 1959;;andArkitektur, from 1960 to 2010.
2. Dining room (sal or matsal).3.Living and bedroom areas.These sometimes included a small living room for the family.Especially in winter, the bedrooms were also important living areas, especially for the woman of the house, who might use the bedroom for receiving female guests, writing, and sewing(Paulsson 1950: 121; Gejvall  1988: 234ff).4. Kitchen area with servants' rooms.The difference between serving spaces and served spaces became more important, as did the effort to keep the sounds and smells of the kitchen and servant's quarters away from the salons and drawing rooms (Lundberg

Figure 1 :
Figure 1: J.G. Destain's plan of Björksund from the 1720s(Selling 1937: 47).Here the room types are still quite generic, relating to the size or position of the rooms rather than to their use, like antechamber (förmak), chamber (kammare) and cabinet (kabinet).

Figure 2 :
Figure 2: E. Palmstedt's plan of Skinnskatteberg from 1770s (from Selling 1937: 324).The wife's quarters are on right side at the the back, whereas the husband's quarters are on the left side at the front.The kitchen (kjök), maid's chamber (jungfrukamare) and a chamber for porcelain (porcelainer) are at the back left side.The dining room (matsal) is connected to the entrance (förstufva), and the great room (sal) is connected to the main stairs on the second floor.

Figure 3 :
Figure 3: Two Stockholm flats, drawn by A. Johansson in 1894-96 (Teknisk tidskrift, Afd.för byggnadskonst, 1897:pl.13).The tambur to the left follows an older tradition where movement through more representational spaces becomes obligatory.The tambur to the right is connected to a set of passages and secondary spaces.Here we can also see a division of the apartment into four different areas, as described above.

Figure 5 :
Figure 5: The death of the salon (salong) and the antechamber/drawing room (förmak) and the rise of the living room (vardagsrum) and the family room (allrum), showing the percentage of residential plans with a specific room type, 1890-1990 (number given for the decade: 1890 comprises 1881-90, etc.).Diagram by Mattias Kärrholm.

Figure 8 :
Figure 8: Allrum with dinner table and a table designated for work and play (with the kitchen in the background); interior design by Lena Larsson.Photo from an apartment in 'Das Schwedenhaus' at the Interbau exhibition in Berlin 1957 (Byggmästaren, 1957: 210).

Figure 9 :
Figure 9: The building of collective laundry rooms in housing areas increased in Sweden during the 1940s.Here is an example from Klippan, built around 1940 (Byggmästaren 1942: 295).

Figure 11 :
Figure 11: Here we can see how the left part of the hall has been designated for play with the inclusion of an early example of the room type lekyta (play area).Row house designed by Gustaf Lettström for the Housing exhibition H55 in Helsingborg (Byggmästaren 1955: 233).

Figure 10 :
Figure 10: The Hovrätt (Court of Appeal) in Malmö, designed by Ivar Callmander (Arkitektur 1919: 139).The ground floor has offices, an archive and an apartment for the building supervisor (on the upper right side).

Figure 12 :
Figure 12: House in Växjö designed by Gösta Brügger (presented in Byggmästaren 1946: 417).Here one can see a small arbetsvrå (study nook) next to the terrace.The house also shows a late example of a borstrum (brush room) next to the entrance.

Figure 13 :
Figure 13: The bodega room is a one-off room type in the journal Byggmästaren(1946: 433), and can be found on the plans for a detached house in Tyresö drawn by Holger Blom and Jan Wahlman.The bodega room was placed in the basement and was intended to be used as a kind of party room.The name did not quite catch on, but the name gillestuga was later used for a very similar room type that played an important role in Swedish dwellings from the early 1960s and up to the 1980s.

Table 1 :
The years of first and last appearances of still-active residential room types, based on the database developed by the author from architectural plans dating between 1750 and 2010.The second column shows the first appearance according to the Swedish national dictionaries (SAOB and SO).The following two columns show the first and last appearances according to the room type database (on which this article is based).The final column shows the age of each room type (in years), calculated as the difference in years between 2010 and its first appearance (marked in bold).

Table 2 :
Lifespan of short-lived or unusual residential room types, based on the database developed by the author from architectural plans dating between 1750 and 2010.The second column shows the first appearance according to the Swedish national dictionaries (SAOB and SO).The following two columns show the first and last appearances according to the room type database (that this article is based on).The final column shows the total number of years since the room type's last appearance in a residential building (in 2010).