How data became the common language of interdisciplinary research

Picture a marine biologist, a linguist, and a software engineer gathered around the same spreadsheet, debating the meaning of a single column header. That scene, awkward as it sounds, has become one of the defining images of modern research. Across the globe, scholars from fields that once kept to themselves are discovering that the currency they trade in is data, and the medium they speak through is increasingly computational.

The themes explored at IASSIST 2017, hosted at the University of Kansas in Lawrence between May 23 and 26 of that year, captured this moment with unusual clarity. Under the banner "Data in the Middle: The Common Language of Research," the gathering examined how numbers, formats, and metadata have become the connective tissue between sciences, humanities, and policy shops. Plenary talks on big data, deep learning, digital agriculture, and global food security showed the conversation had moved well beyond any single department.

For researchers based in Australia, the same discussions unfolded with their own cadence. Brisbane, Melbourne, and Hobart hosted major data-focused gatherings in the years around the conference, and institutions such as CSIRO, the Australian Research Data Commons, and several Group of Eight universities began investing seriously in research data infrastructure. The country became an active participant in a wider research culture that already shared its vocabulary across the Tasman and the Pacific.

The origins of cross-disciplinary data work

Long before "big data" became a boardroom buzzword, scholars had collaborated across disciplinary lines for decades. The difference is that earlier collaborations relied on personal relationships, shared fieldwork, or a senior professor's network. Data was an output, sometimes an afterthought, often locked inside a filing cabinet or personal hard drive. When projects ended, the information frequently vanished with them.

By the late 2000s, that model had begun to creak. Funders from the National Institutes of Health to the European Commission demanded data management plans, and journals tightened reproducibility requirements. Australian researchers felt the same shift, with the Australian Research Council introducing data-sharing expectations and the National Collaborative Research Infrastructure Strategy funding platforms such as the Atlas of Living Australia. Conversations that once happened in corridors moved into standardised repositories.

What changed the conversation most profoundly was the realisation that data, when properly described, could outlive its original project. A drought-monitoring dataset collected by a hydrologist in the Murray-Darling Basin could feed a policy paper, a master's thesis, and a museum exhibition. That possibility, of data becoming a shared resource rather than a private archive, is the foundation of the interdisciplinary movement.

Why data became a shared vocabulary across fields

Every discipline develops its own jargon, and researchers learn to translate when they cross into another field. Data, paradoxically, offered something better than translation: a lingua franca that did not depend on the original language of the question. A raster image of a coral reef, a CSV file of wool prices, and a corpus of transcribed oral histories can all be described using the same metadata standards, even when the underlying methodologies differ wildly.

That technical common ground was reinforced by a cultural shift in training. Doctoral candidates in Adelaide, Perth, and beyond now routinely take courses in research data management, version control, and reproducible workflows. Supervisors who once told students to "keep good notes" now point them toward GitHub, Zenodo, and ORCID. The result is a generation of researchers who treat data work as a craft, not a chore.

Standards such as Dublin Core, DataCite, and the FAIR principles gave the field a shared grammar. Australian contributors played a meaningful role in shaping these standards, particularly through the Research Data Alliance and the development of Australian and New Zealand research data management guidelines. The result is a vocabulary that allows a soil scientist in Wagga Wagga and a digital humanities scholar in Dunedin to discuss the same dataset without first negotiating terms.

Big data and the rise of deep learning in research workflows

When the Lawrence program put big data and deep learning alongside digital agriculture, it signalled something important: machine learning had stopped being a niche specialty and started becoming a research utility. Pattern recognition, once the domain of statisticians, became accessible to anyone willing to spend a weekend with a tutorial and a Jupyter notebook.

The Australian uptake has been striking. CSIRO's Data61, headquartered in Sydney and Brisbane, has built national capability in machine learning, computer vision, and autonomous systems. The Murchison region of Western Australia hosts the low-frequency antennas of the Square Kilometre Array, producing volumes of astronomical data that have forced researchers to rethink storage and processing. Regional universities have embedded data science electives into undergraduate agriculture and environmental science degrees.

Practical workshops at IASSIST meetings became a way for newcomers to find their feet, with topics ranging from data cleaning in OpenRefine to introductory deep learning with TensorFlow. Participants left with more than new technical skills; they left with a clearer sense of how their research questions could fit into a much larger methodological conversation.

Digital agriculture and global food security on a warming planet

Few areas illustrate the interdisciplinary power of data better than agriculture. Crop modelling, soil chemistry, climate science, economics, and supply-chain logistics all feed into the question of how the world will feed itself in 2050. Without a shared data infrastructure, those conversations would stall in the first hour. With one, they can run for years.

Australia's experience has given the country a particular voice in these discussions. Decades of drought, salinity research, and water reform have produced rich datasets used by international collaborators. The Grains Research and Development Corporation funds research that links farm-level yield data to satellite imagery and climate projections. In a country where a bad season can mean the difference between profit and ruin, the analytical stakes are personal as well as scientific.

The digital agriculture strand of IASSIST 2017 brought together agronomists, data engineers, and policy analysts. Sessions highlighted open-source tools for precision agriculture, frameworks for sharing farm data ethically, and case studies from sub-Saharan Africa, South Asia, and Australia. The recurring theme was that data is necessary but not sufficient: it must be paired with local knowledge, community trust, and clear governance.

Skills, literacy, and the rise of the data practitioner

A new professional class has emerged around research data, and it has yet to settle on a single name. Some institutions call them data stewards, others research technologists, and still others digital research consultants. Whatever the label, the role is the same: to help researchers across a university handle their data responsibly, from planning through archiving.

Typical responsibilities in the role include:

• Drafting institutional data management policies and templates
• Running training sessions on research data management tools
• Liaising with IT services on storage, backup, and access controls
• Advising grant applicants on data management plan requirements
• Supporting researchers through ethics and privacy review processes
• Tracking compliance with funder and journal data policies

The shape of this role varies by institution. At the University of Melbourne, data professionals are often embedded within faculties, working alongside historians and biologists in equal measure. At the Australian National University in Canberra, dedicated teams support large-scale infrastructure projects, including the heavy data demands of the ANU Bioinformatics Consultancy. In regional centres, a single data specialist might serve an entire campus, juggling training, consultation, and infrastructure support in the same week.

This distributed model has consequences for career pathways. There is no single qualification that prepares someone for the role, and practitioners often arrive from library science, statistics, software engineering, or the disciplinary sciences themselves. The Australian Library and Information Association, eResearch Australasia, and several state-based research offices now offer professional development streams that recognise this hybrid identity.

Conferences as crossroads for cross-pollination

It is worth pausing on the role conferences play, because they often get written off as mere networking opportunities. In the data community, they serve a more specific function: they are where the shared vocabulary gets rehearsed, challenged, and updated. A taxonomy that works in the lab may not survive contact with a legal scholar or an archivist, and conferences are where those collisions happen in person.

IASSIST 2017 in Lawrence was a textbook example. Sessions were deliberately cross-disciplinary, with panels mixing data librarians, domain scientists, and software developers. Plenary speakers, drawn from organisations such as Microsoft Research, the Consultative Group on International Agricultural Research, and several leading US universities, framed data work as fundamentally collaborative. Australian attendees, often clustered around posters on the Atlas of Living Australia or the Terrestrial Ecosystem Research Network, found their work recognised in a global context.

The pre-conference workshops schedule from that year remains a useful artefact. It listed sessions on data carpentry, social science data archives, and reproducible publishing, giving newcomers a structured on-ramp. Many of those workshops have since evolved into recurring online training events, ensuring that the on-ramp stays open long after the closing reception ends.

Sustaining the conversation beyond Lawrence

Conferences end, but the work they seed continues. The threads that ran through IASSIST 2017 have since woven into new initiatives, new policies, and new community groups. In Australia, that has meant deeper investment in the Australian Research Data Commons, the launch of new data skills training through the Australian Technology Network, and continued advocacy for Indigenous data sovereignty through bodies such as the Maiam nayri Wingara Aboriginal and Torres Strait Islander Data Sovereignty Collective.

The broader lesson is that data is not, by itself, a solution. It is a medium through which solutions become possible. Researchers who treat it as a first-class research output, rather than a byproduct, tend to produce work that travels further and lasts longer. That shift, more than any particular tool or standard, defines the interdisciplinary era.

Australian research groups have adopted several practical habits that could be useful elsewhere:

• Writing data management plans at the proposal stage, not after the fieldwork is done
• Pairing every PhD student with a data steward for at least the first year
• Publishing negative results alongside positive ones, especially in climate and ecology work
• Linking datasets to publications using persistent identifiers such as DOIs
• Consulting with Aboriginal and Torres Strait Islander communities before reusing cultural data
• Reviewing metadata quality annually, even for established repositories

The differences between disciplines, countries, and institutions are real and worth honouring. Yet the comparison below shows how the underlying data principles travel surprisingly well:

Aspect	Library and archive tradition	Scientific computing tradition	Domain-specific research tradition
Primary concern	Long-term preservation	Reproducible computation	Domain accuracy and relevance
Typical tooling	DSpace, EPrints, Archivematica	Git, Jupyter, Snakemake	R, SPSS, discipline-specific platforms
Metadata emphasis	Bibliographic and provenance	Computational environment and versions	Variables, instruments, sampling methods
Career pathway	Library science qualifications	Computer science or statistics degrees	Discipline-specific PhDs with added skills
Engagement with policy	Strong	Growing	Variable

Different starting points, similar destinations. The Australian experience, from the mallee scrub of South Australia to the data centres of Canberra, illustrates the point. When researchers agree on the basics of how data is described, stored, and shared, the rest of the conversation becomes far more productive. That shared foundation is the real common language, and it is the one IASSIST 2017 set out to articulate.