Whenever you begin a new research project, you need to think about the methodology you want to employ in order to collect and process data, produce and communicate results. In many cases this basically means “Hey, let’s use a mobile application for data collection because it looks cool and do everything else like we’ve done it before because it’s convenient”. And to be honest, there are a lot of good reasons to plan a study like that. Facing the unknown is never easy, but especially when you depend on external funding, just entering “Look, the idea is neat, so let’s just see where it gets us” might not be the best option, at least if you want to stay in Academia.
So when I came up with the idea for my new project Lingscape, maybe I was just lucky three times over: not to depend on external funding, having open-minded colleagues to plan and implement this project with, and being driven by a general interest in the theoretical foundations of research methodology and the consequences thereof for fieldwork. Be it as it may, I ended up asking myself and others a lot of questions regarding my idea, the methodology to apply, and how this would impact the whole research process as well as my own role as a scientist in this connection.
But let’s start with the idea: Working in a multilingual society (and university) in Luxembourg, the manifold ways in which the presence and absence of certain manners of speaking (or “languages” if you will) in public signage communicates the social constitution of a place or society quickly caught my attention, especially against the background of a flustered discourse in the media about the role of Luxembourgish as a national language and major means of sociocultural ideologization. Of course I had heard about the rise of research on the semiotics of signs and lettering in public space under the label “linguistic landscaping”. But taking a closer look at many of the available (mainstream) studies in this vital subbranch of Sociolinguistics, I kept wondering why these often insightful studies hadn’t tapped into smartphone technology and crowdsourcing for their research, not only to gather way more data than a small research team normally could, but mainly to move beyond our scientific perspectivation of semiotics and cultural practice that in a way is limited by our professionalization.
To cut the story short: I wanted my own linguistic landscapes project, but I wanted it to be a Citizen Science project, so the solution was “Hey, let’s use a mobile application because it looks cool…”. I started to look out for partners in my institute and a software studio that would help us with app programming, and so the journey began. Now, after a year of developing, testing and advertising we have published the app under the name Lingscape and shaped the project along the slogan “Citizen science meets linguistic landscaping”. You can easily find out more about the project here. The app looks nice, works smoothly, and there’s a lot of promising directions to tap into with it. For example we are cooperating with the Luxembourgish Ministry of Education to test and implement Lingscape as a digital teaching tool in primary schools. So far so good. And of course I have written a first paper about the background of the project, the theoretical demands for a picture-based approach to crowdsourcing and the methodological guidelines for both app development and the pilot study in Luxembourg. If you’re interested in that, you can find the paper here.
But now, only six weeks into data collection, I find myself confronted with the various methodological implications to this citizen science approach that one needs to take into close consideration in order to implement this research properly. Because doing research like we want to do it with Lingscape brings about some substantial changes: First, it alters the data base for our studies in replacing our limited professional perspective on phenomena, e.g. signage in public space, with a multitude of individual perspectivations thereof provided by users. And second, it dramatically impacts all phases of the research process as well as our role as scientists. Plus, the scope and potential impact of the project unfold in domains and directions that I hadn’t foreseen in full when coming up with the idea for the app. Turns out the project is way more about societal dynamics, participation and the manifold ways in which social actors negotiate sovereignty of interpretation over public space than just about counting languages on public signs and putting them on colorful maps. I guess that’s not the worst thing to happen, if a project exceeds your expectations that quickly while thoroughly challenging the way you see yourself as a scientist.
So picking up the challenge of reflecting these changes and their impact on my research at the outset of this new and exciting project, here’s what I consider to be the methodological groundings of a sustainable community-driven Lingscape research. Please note that although all bullet points may say “… Science”, these are not supposed to be proper or even new disciplines; they are just constitutive aspects of a comprehensive scientific practice that essentially contribute to the way I understand the Lingscape project, well, now:
- Citizen Science: Data collection with Lingscape revolves around the users' contributions and perspectivations of the linguistic landscape that surrounds them. This is what Citizen Science mainly is about: engaging citizens with scientists in a joint research process. But taking user engagement seriously does not only mean to value their contributions as data suppliers. It also means to give the users a share in the development of the project as well as in data handling, analysis and the exploitation of results. Therefore we have been including users in the development of the mobile app and the scope of our research right from the start of the project to allow for a common perspective on how this project can contribute to both academic research and societal practice. Citizen Science is research with society, not about it.
- Embedded Science: Taking user engagement seriously also requires the direct embedding of research in the users' everyday life, not the other way around. Often academic research approaches cultural practice by creating experimental environments by which specific aspects of human action shall be monitored and surveyed. While this might be useful in many ways, we want to take a different approach with Lingscape, namely directly embedding our research tools in our users' social routines. Therefore we are for example cooperating with the Luxembourgish Ministry of Education to implement Lingscape as an official teaching tool in elementary schools that will help to sensibilize pupils for cultural diversity in their lifeworld. Working with teachers and children allows us to shape our research based on the pupils' experiences with the app in everyday life, which leads to better teaching tools and an embedded scientific practice. Embedded Science is research in society, not above it.
- Engaged Science: If we want our research to directly reflect social dynamics in society, we also need to engage directly with the communities we want to survey. This basically means avoiding the reduction of social practice to a set of phenomena and variables for analysis so that it becomes possible to develop shared interest with citizens and stakeholders in order to implement our research according to the communities' interests and needs. Basing the project on the specific social practice of a community allows for the fostering of sustainable cooperation that helps to improve the way communities organize and develop their social practice. In fact one of the main ideas for Lingscape is to help redefining the relationship between scientists and society. We want to work not only with and in society, but for the people who enable us to do our research. Engaged Science is research for society, not because of it.
- Computer Science: Crowdsourcing as a method surely is nice because it allows for huge collections of photos representing a multitude of different lingscape perspectivations. But analyzing these collections also requires new approaches to data analysis. Regarding photo-based lingscape research this includes a lot of advanced computational methods that come in pretty handy for the analysis of huge photo corpora as well as the visualization thereof, for example automated text extraction and language identification from photos as well as tools from natural language processing for data handling, analysis and mapping. Using such tools and methods for our research does not only facilitate the efficient processing of huge data collections but also helps to open up complete new domains of knowledge creation in linguistic landscapes research compared to the more traditional “walk-snap-count” approaches. Computer Science is a means to a research end, not a research end in itself.
- Open Science: With all that said, of course all research we pursue within the Lingscape project will be open. This does not only affect the availability of data sources (as natural as indispensable for a Citizen Science project) but also the accessibility of results and other exploits from our studies, be they of academic or popular kind. Moreover, open science requires transparency regarding all processes of data gathering, handling and evaluating so that not only the foundation of scientific knowledge production, aka the data, is open to society but the whole structure of interpretation that makes research. This might not only help to close the gap between society and academic research that sometimes may seem a bit opaque and remote from everyday life to the outsider. It is also both an invitation and motivation for us researchers to constantly reevaluate and develop further the theoretical and methodological foundations of our work. Open Science is research that regards transparency as a chance to its impact on society, not as a challenge to its sovereignty of interpretation over society.
In short, my (current) idea of what we are trying to do in the Lingscape project is research that is driven by citizens, embedded in and engaged with society while making use of cutting edge computational methods of analysis to collect open access data and establish a transparent research practice. I believe that such a comprehensive approach to linguistic landscape research allows us to survey the cultural complexity of communities together with the people in a sustainable manner. But it could also help us to develop and implement innovative research that matters to society. So my hope here is that Lingscape may be just the perfect tool to tell this different kind of science story.