“Alive and kicking” – Moving through and diving into the Soviet kolkhoz and the East German LPG

. The MODSCAPES project seeks an understanding of how modernist rural landscapes developed over time, in part calling for an understanding of the landscape as a set of intertwining layers assessed by local mapping. In addition, by understanding the spatial grammar of the landscape as well as perceiving it through different media, other aspects can be revealed which are not visible via mapping alone. The East-German and Baltic cases differ from other European examples by the fact that their existence ended nearly 30 years ago and residents and decision-makers from the time are still alive. Thus, we focused on the importance of actions carried out by residents in their everyday lives and ways to connect them with the respective space. If a landscape is understood by layers, then these actions form the “kebab skewers” metaphorically connecting them. Topographic maps from different periods formed the basis for the experiential data collected and interpreted in related steps accompanied by verbal commentaries. Firstly, we “dived” into the area using filming and field recordings simultaneously like a canvas to paint on and in the sense of a journey. Next, 360° surround films were shot at spots to simulate the view of a person turning around, followed by filming of situations representing everyday movement cycles in the area, such as going to work or taking children to school, which evoked an atmosphere of everyday life linking the space and people’s actions. Finally, go-along interviews were used to trigger and stimulate reflections and memories of residents to understand how the space impacted their experiences and perceptions. This process revealed facets of the daily life of the inhabitants, settlers or workers and their social interaction with the landscape, uncovering so far untouched places and unknown spatial relationships.


Introduction
In the MODSCAPES project, the team at the Estonian University of Life Sciences are studying the phenomenon of agricultural collectivisation in Estonia, Latvia and the former GDR. We are interested in understanding the relationship between the landscape of collectivisation at different scales (the wider landscape and the modernist village) and the experiences and perception of residents who lived (or live) and worked there, potentially before, during and after collectivisation. In this paper we present and discuss a number of visual mapping techniques we are using too assess three of our MODSCAPES case study areas located at Omedu and Laeva in Estonia, Triinwillershagen in Germany and Seda in Latvia (contrasting examples within the wider case study assemblage of the MODSCAPES project). The research objective was to explore how the landscape in these areas developed, what it is like now, what is still visible and what is its condition as perceived, experienced and endured by the inhabitants. Various data gathering methods from other research fields were reviewed and a combination was formed into a methodology for testing [1]. By understanding the spatial grammar of the landscape as well as perceiving it through different media, aspects can potentially be revealed which are not visible via visual and structural mapping alone, but by connecting these characteristics with the active life of people. This paper therefore combines a discussion of methodological development with presentation of results for the three cases.
The Baltic and East-German examples differ from other European MODSCAPES case study areas by the fact that their existence ended only 30 years ago and residents and decision-makers from that time are still alive (and kicking). The collectivisation of agriculture in the Baltics and the GDR started in around 1949 and ended with the collapse of the Soviet Union in 1991. The process turned property owning peasants into salaried workers with regular working hours and social security. At the same time specialized agricultural engineers were educated to manage the new collective farms, usually incorporating several farmsteads or even villages. These collective farms were in need for planning and constructing extensive field drainage systems, paved roads and residential buildings for the farm workers, propelling the remote rural cultural landscape into what we define as a MODSCAPE.
After gaining independence in the Baltic States and the reunification in Germany 30-25 years ago, the land was either given back to former owners or sold, which completely questioned and transformed the whole economic, social and spatial structure of the collective farms into their present state. The buildings with a certain cultural value have sometimes undergone restoration but the Modernist, socialist settlement structure is frequently neglected. However, the villages and their surrounding cultural landscape are part of the ideological cultural landscape heritage of Eastern Europe and the migration of people and cultural decline of these very extensive areas pose a challenge in terms of how to deal with this layer of history, such as how much to preserve, as well as raising awareness about this heritage. To emphasize that these places are actually alive and kicking and not abandoned or museums, different data capture techniques can be applied and combined.

Places
Landscapes where people live become places and are experienced not (only) from static viewpoints (as scenery) but by moving within and around them. Since landscape is "…an areas as perceived by people…") [2], a visual approach through movement (such as walking or driving) to explore the representations of the landscape as experienced by their people is appropriate, because "the aspect of walking allows for an understanding of places being created by routes" [3: 244]. A sense of place is here captured through the through the use of audio-visual media following a representative sample of routes. Audio-visual representations offer the possibility to place ourselves, as researchers, into these landscapes and to understand the place-making practices directly in order to understand the emplacement of inhabitants, which the accompanying mapping process enables people, experiences and space to be understood together.
Thus, we might walk or drive through an area while talking to some inhabitants in order to trigger their responses and memories of specific places. Inhabitants are prompted by their connection to the place and talk about it while being in that place.

Methods
The methodology presented here is built from a critical review of existing research methods but developed for the MODSCAPES research. In the overall research itinerary, we examine the pattern and structure of the landscapes from different periods using historical GIS [4] forming the basis for the experiential data. For each case study area maps of the area before, during and after the collectivisation period were obtained at different scales and periods, created by using different map projections converted into vector GIS files as a base for localizing the filming steps and as a trigger during the later interview steps. We wanted to collect and interpret moving images made by videos while driving and walking through or around the case study areas combined with verbal commentaries, while interviews with residents and former workers revealed the way in which the territory operated and what each area was for.
The first step focused on the space in general being observed by outsiders and the second step took a resident's perspective; but to understand how the landscape impacted and still impacts the experiences and perceptions of residents both stepst need to be connected. Thus, we then focused on the local residents in their present and past everyday lives and how they were connected to a specific place. If a cultural landscape is understood as a series of layers, then the past and present everyday lives of inhabitants form some kind of "kebab skewers" metaphorically connecting the earlier layers with the present to give us a more complete understanding of place and, furthermore, to create public awareness of them as an output of the project.

The go-along method
Go-along interviews were developed and applied first in ethnological research, followed by health and traffic research. According to Evans and Jones Go-along interviews are s a qualitative method for studying the health issues of a neighbourhood context in which people live [5]. Carpiano notes that his respondents found it easier to verbalise attitudes and feelings when being in place [6]. It is an interviewing technique that explores peoples' experiences in their familiar environment. Kusenbach has offered the most advanced discussion on the go-along method, assessing the role of place in social problems from an ethnographic perspective [7]. She identified five themes for which the Go-along is suited: perception, spatial practices, linkages between biography and place, the social architecture of natural settings and social realms -all relevant for this project.
The route for the Go-along interview is either set by the interviewer or the interviewee, but according to Jones et al. determining a set route has the advantage of focusing the interview on specific places that are relevant to the goals of the research project [8]. If the researcher wants opinions about specific buildings or paths this might require following a structured route, but some routes can be determined by the interviewee made up on the spot according to the contextual trigger.
Locating the discussions on a map with markers adds another layer of interpretation to the data collected, since one of the goals is to consider context, allowing spoken words and spatial location to be directly connected. However, only focusing on words and location means that the main characteristics of the Go-along interview can be lost. By using video the necessary visual context is provided, although filming; walking and talking at the same time can be difficult. Nevertheless, for our research we tested whether a comprehensive Go-along interview approach with video could gather more information about the landscape than by using interviews and observations separately. As mentioned above, a filmed representation captures a place through movement, voices and/or visual narratives, meaning that the videographer needs to learn where to look. Whether a filmed interview can serve as a connective trigger for gathering information needed to be tested. This test could be carried out by comparing the output of the Go-along approach with other more traditional ways of interviewing. Go-along interviews were carried out using a semi-structured format and used prepared questions from the MODSCAPES research protocol, as well as a list of context-related pre-prepared questions/topics to discuss. In this way, the interviewer only gave the interviewees some direction, leaving them free to comment on what they saw. It left enough time for the interviewer to point out features or elements in the area, giving a deeper insight by the interviewee reflecting on it and turning the camera from the interviewee to the respective element (Fig. 1). The Go-along interview connects field observation with traditional interviewing, taking "… advantage of each method strengths, while employing both to compensate for each other's limitations" [6 : 265]. Although the Go-along approach is characterised as being a Jack-of-all-trades for mapping and understanding a spatial context, it needs to be connected with preparational steps, not necessarily connected to each other from the start. Several steps were used, each building on the last (Fig. 2): the field observation is for first-hand familiarisation with the location, being useful for assessing or mapping features of the social and physical environment but "…limited by the researcher's own interpretive framework." [6: 266], while traditional interviews have limitations for examining experiences at a place because participants won't discuss issues or ideas for which they are not immediately aware [9]. As a first step, we "dived" into the case study areas using filming and field recordings simultaneously, like travelling through the landscape to narrate a story.

Results
In this section examples of each step are presented using different examples in order to show how they work under different conditions.
The Dive-In step we researchers moved around on foot to perceive a small-scale building and its surroundings first-hand. This was conducted through filming and adding our own comments about route and area, creating a continuous space through random movement for a short period of time. As an example, the first encounter into Omedu, an abandoned factory area in Estonia (Fig. 3). built the basis for the later steps in the research narrative. As a quick but intensive recording phase it gave a first understanding of the spatial character as a canvas to paint on. The material was used for assessing types of distinctive places and elements being captured multiple times marked in the GIS maps. In the Drive-Through step we, a team of two, drove around the larger-scale former socialist model farm of Triinwillershagen in Eastern Germany by car (Fig. 4) while filming and recording comments about route and area. It created a sense of the continuous space at a large scale, creating a sequence-like encounter with fields, meadows and rows of houses, unfolding through discussions in the car. We used a stabilized action camera to film at a familiar height and ensuring smooth movement without focusing. The film material was then used to assessing types of distinctive places or elements which were filmed multiple times. The route mapping and important place locations were added to the GIS map later. The Look-Around step was used in three study areas (Omedu, Triinwillershagen and Laeva) to explore their character using 360° surround filming with recorded commentary at specific spots to be shown to interviewees in later steps (Fig. 5). These spots were identified during the first two steps after marking them on the map but could be adjusted or added after obtaining information later. The look around simulated the view of a person turning their head around at the spot as if filmed on a tripod. The Look-Inside/Look-Outside step explored two areas by hand-held filming of situations representing everyday movement cycles such as going to work (Omedu) (Fig. 6) or taking children to school (Seda, a Latvian Stalinistic settlement). These locations were also identified during the first two steps but could also be adjusted or added later. To look inside means to invoke an atmosphere of modernist life. By linking space and action, the filmed material will also be shown later during interviews to trigger information of rural everyday life and to identify neighbourhood-gathering places etc. The final step of data collection involved d Talk-about and Go-along interviews, as well as group discussions (with master students) about the results that the steps have revealed. These were important for understanding how the landscape impacted and continues to impact the experiences and perceptions of today's residents and former workers. If possible, they were performed as video interviews with consent of the interviewees.
In the Talk-About in Triinwillershagen and Laeva we sat with the interviewees around a simple map and a computer pad in order to trigger and collect information about places of interest identified by the interviewees on the map. During the interview we showed pictures and films taken in the first steps and asked questions if or how the interviewee related to them (Fig. 7). We pointed out the places displayed on the map and the interviewees described their spatial relationships on the map, while we filmed with the camera focusing on the map. At the Go-Along step we simultaneously collected information about the areas and their actors to explore the relationship between both now and in the past. As mentioned in the methods section, the Go-Along is dependent on continuous movement over a period of time spent with the interviewee while conducting the interview. During the interview, we created a social interaction, building up trust for further encounters and interviews and also triggered interviewees interactions with other people on the street. We also uncovered larger spatial relationships in the area, which would have been impossible in separated observation and interview steps.
In Omedu, the Go-Along interview was clearly the culmination of the image and action data gathering process, while being completely reliant on the previous steps and the GIS maps. We accompanied the interviewees around different places, while then describing what mattered to them personally at the time of collectivisation, asking them questions about the area and how they were related to it. This gave us access to facets of daily life and routines of interviewees as inhabitants, former settlers or former workers. The Go-along was performed on foot as well as by car to capture the information in the first step scales, but also to keep the human scale and the natural proximity between interviewer and interviewee (Fig. 8). We filmed the interview while walking or driving using two researchers: one interviewing, and one filming. Both the interviewee and the traversed space were filmed while interviewing using a small action camera, ensuring that the interviewee did not feel self-conscious and to be able to move the focus easily onto the space the interviewee was describing. In Triinwillershagen the physical health of some interviewees prevented them from walking with us so we talked to them in our car, conducting a drive-along interview to solve the problem.
We always returned to the site of our initial conversation. The interview route was mapped afterwards. The most mentioned, most discussed spots from the interview were added to the GIS maps. These formed the basis for further Go-alongs on this route with other interviewees. After this phase the film material will be further analysed by searching the soundtrack for keywords and assessing types of distinctive places traversed during the interview multiple times.
All steps were also performed simultaneously at the third location, Seda, where we used master students for conducting the filming, commenting and interviewing, which developed into a Scene-Talk discussion. The students made handwritten notations on a map with specific locations or landmarks encountered during the process. By merging the outcomes of earlier steps, the Scene-Talk step triggered a discussion on site among the student group while simultaneously pan-shot filming places the group had encountered there (Fig. 9). The film and recording material will be used for exhibition purposes in the dissemination phase.

Discussion
The mapping steps produced a significant amount of data recorded on the GIS maps as well as image and film footage with the Go-along interviews specifically producing time-stamped interview transcripts and GPS tracks. The analysis involved mapping the whole Go-along interview in a GIS map through the creation of spatial transcripts. Speech objects will be used as units in the text analysis, capturing characteristics for each speech object, such as the type of place, whether it still exists and if it changed its function. The methods we have developed and tested seem able to allow us to examine the important relationship between location and speech objects and also to explore differences between data generated by walking and traditional interviews in some case study areas.
This method, or series of methods, is based on existing approaches but also makes use of developing technology such as lightweight self-levelling video cameras to capture data. The interactive nature of the methods combines what people say with where they say it and about what they say, making closer relationships to the wider landscape possible. When related to methods such as historical mapping even further associations can be made. This demonstrates the benefits of multi-methods for the kind of research applied here.

Conclusions
We aimed to explore and test the possibilities of different combined methods for capturing the place relationships over time of people and their landscape. The steps culminating into the Go-along interviews led us along on a mapping journey through the MODSCAPES case study areas. We learned about the areas by means of making explicit the relationship between interviewees' ideas or attitudes and our own experience in the respective environments. The interviewees became more of an equal participant in the interview than only a subject who is being interviewed. All interviewees did not rush through the interview process or ended the interview earlier. There were almost no pauses in the conversation and the interviewees actually enjoyed telling their life story to us. Hence, Go-along interviews were poking the interviewees as residents and former farm staff to develop pride and awareness about their former collective landscape as it has been shaping their active work life.
Go-along interviewees liked talking about specific buildings or landscape features while the traditional interviews produced narratives that focused on people. This shifted our mapping method towards a combined approach, where Go-alongs are connected to field observations and traditional interviews in a well-directed process. Go-along interviews produce a discourse on different places. Mapping these interviews produces a narrative that only unfolds through place. By using GIS maps, it was possible to incorporate narratives from the interviews together with quantitative spatial data. Goalongs can be incorporated with student groups but also focus groups to generate a discussion in a location and to communicate opinions or ideas regarding possible development issues.