Whose Reality?

Unpacking data in Architecture.

Data has rapidly evolved from a background technical tool into one of the most influential elements shaping contemporary architecture. Yet, despite its prevalence, architects rarely question what ‘data’ fundamentally represents or acknowledge how deeply it influences their work. This post addresses that gap by exploring data’s character and implications for architectural design—from initial observation and analysis to creative forecasting and eventual material realisation.

Data as Raw Material: Setting the Scene

The concept of data entered the English lexicon in the seventeenth century, derived from the Latin ‘datum’, meaning ‘something given’—an unquestionable starting point (Rosenberg, 2013). During the Enlightenment, however, the term underwent a dramatic shift. No longer an unquestionable truth, data became associated with empirical observation and measurement, fostering trust in phenomena that could be scientifically observed and recorded. The Royal Society of London’s motto, “nullius in verba” (take nobody’s word for it), encapsulated this shift toward evidence-based, mechanical observation, profoundly reshaping how knowledge was produced (King & Kay, 2020).

Yet this transition to empirical observation introduced new limitations. Human senses, although fundamental, are inherently subjective and limited—able to perceive sound, colour, and spatial qualities, yet incapable of detecting invisible phenomena such as electromagnetic fields. Consequently, data captured through human senses inevitably present only a partial view of reality (Harman, 2011).

The advent of mechanical observation technologies, such as the microscope and photographic cameras, further shifted cultural attitudes towards observation. Luciano Floridi (2014) refers to this phenomenon as "mechanical trust", a belief in the superior accuracy of precision instruments over human senses. Etienne Jules Marey’s photographic studies of motion exemplify this, capturing details invisible to the human eye, cementing reliance on machine-generated data (Giedion, 1948).

Today, this mechanical trust is amplified through digital tools, automated systems, and advanced sensors widely adopted by architects. However, architects risk implicitly accepting machine precision without critically examining underlying assumptions or potential biases embedded within these methods. Rob Kitchin emphasises this risk, defining data as “the raw material produced by abstracting the world into representational forms” (Kitchin, 2014, p. 2). Thus, architects must critically examine how data transforms their perception and practice, considering how hidden biases could impact their designs.

Data Identity: The Role of Meaning

Understanding data’s role in architectural practice demands attention not only to how it is captured but also how meaning emerges from it. Russell Ackoff’s Data-Information-Knowledge-Wisdom (DIKW) pyramid illustrates data as symbolic representations awaiting human interpretation to become meaningful (Ackoff, 1989). Ackoff argues that data gains value when patterns emerge, answering specific questions.

In contrast, Jennifer Rowley (2007) challenges Ackoff’s hierarchy, proposing an inverted model where data sits at the top, containing all potential meanings. For Rowley, data continually exists as a resource, from which information, knowledge, and wisdom are distilled through progressive ordering and interpretation. The distinction between Ackoff and Rowley highlights two critical modes of thinking: data as a foundation awaiting meaning, or data as a resource continuously refined into meaning.

Luciano Floridi (2017) further develops the understanding of data’s identity by introducing ‘dedomena’—pure data existing independently before interpretation. Floridi clarifies that both human and machine data result from selectively capturing phenomena from reality. This implies data is never neutral; it always embodies cultural assumptions regarding objectivity, precision, and purpose, which architects must consciously acknowledge in their practice.

Mapping as Cultural Interpretation

Acknowledging data’s constructed nature leads naturally to understanding how architects assign meaning through representation. James Corner (1999) emphasises mapping as inherently selective, interpretive, and culturally embedded. Maps do not simply reproduce reality; they selectively arrange differences—topographic, political, or spatial—to communicate specific meanings or agendas.

Cartographic practices historically reveal this selective power, embedding biases through choices of representation that influence perceptions of reality. Architectural drawings and visualisations operate similarly, selecting and emphasising specific data points while omitting others. This selective visual language shapes not only the viewer’s perception but also constrains or expands the architect’s creative possibilities.

Thus, mapping serves as a powerful metaphor for architects, illuminating how choices in representing data significantly impact design outcomes. Understanding this relationship helps architects recognise and question implicit assumptions embedded within their own mapping and visualisation practices, ensuring more informed and responsible design decisions.

Forecasting the Future: From Observation to Proposition

Mapping establishes a critical bridge between observing existing conditions and proposing future architectural realities. Rather than relying solely on imaginative processes, as Michael Hays (2019) suggests, architects actively forecast potential futures by interpreting patterns from observed data. Forecasting explicitly recognises architects’ predictive claims, highlighting the responsibility embedded within their creative decisions.

However, forecasting becomes particularly complex when architects unquestioningly trust synthetic, algorithmically generated data. Increasingly prevalent in predictive models—such as energy performance simulations or occupancy analytics—synthetic data inherently carry biases embedded in original datasets, potentially presenting a false sense of neutrality and certainty (Rosenberg, 2013). For architects, this requires a heightened critical awareness. Rather than passively accepting predictions, architects must question and challenge these assumptions, recognising synthetic data’s potential to reinforce existing biases or to restrict rather than expand architectural possibilities.

Realising Ideas: Instruction and Materialisation

The translation of predictive design proposals into built forms introduces another critical interaction with data: instruction. Robin Evans (Evans & Mostafavi, 1997) identifies the "space of translation"—the interpretive gap between architects’ conceptual intentions and their physical realisation. In practice, architects rarely build directly; instead, they provide instructions through drawings, specifications, and digital models, which others interpret during construction.

This translation further compounds partiality, especially with increasing reliance on digital fabrication, automation, and robotic assembly. Although these technologies narrow the gap between design intention and material realisation, they risk prioritising measurable precision over the adaptability and interpretive flexibility inherent in human construction practices.

Post-occupancy evaluation (POE) exemplifies this complexity. Intended to hold architects accountable to real-world performance criteria, POE inadvertently shifts architectural success toward quantifiable metrics—often defined externally by clients or economic imperatives—potentially constraining creative innovation and experiential richness (Preiser, White & Rabinowitz, 1988).

Implications for Architects Today

As architects enter an era profoundly shaped by data, the need for critical engagement intensifies. Firstly, architects must maintain vigilant awareness of partiality, critically interrogating assumptions and biases embedded in their methods of data collection, analysis, and interpretation. Ignoring this step risks unintentionally encoding these biases into designs under the false guise of objectivity.

Building upon this critical foundation, architects must skilfully navigate synthetic predictions produced by increasingly prevalent computational models. They must resist passively accepting algorithmic authority, instead challenging these models to ensure they expand, rather than limit, creative exploration.

Finally, architects must carefully manage the translation from data-driven design to built realities. While embracing technological precision, they must preserve space for human intuition, interpretation, and adaptability, ensuring designs remain culturally sensitive and experientially rich.

By actively addressing these implications, architects can harness data’s powerful potential while maintaining their creative autonomy and ethical responsibility, guiding architecture thoughtfully through the digital age.

Ackoff, R. L. (1989). Journal of Applied Systems Analysis, 16, 3-6.

Corner, J. (1999). Mappings. Reaktion Books.

Daston, L., & Galison, P. (2010). Objectivity. Zone Books.

Evans, R., & Mostafavi, M. (1997). Translations from Drawing to Building. MIT Press.

Floridi, L. (2017). Semantic conceptions of information. Stanford Encyclopedia of Philosophy.

Giedion, S. (1948). Mechanization Takes Command. Oxford University Press.

Harman, G. (2011). The Quadruple Object. Zero Books.

King, M., & Kay, J. (2020). Radical Uncertainty. Little, Brown Book Group.

Kitchin, R. (2014). The Data Revolution. Sage.

Preiser, W. F. E., White, E., & Rabinowitz, H. (1988). Post-Occupancy Evaluation. Taylor & Francis.

Rosenberg, D. (2013). Data before the fact. In L. Gitelman & V. Jackson (Eds.), Raw Data is an Oxymoron. MIT Press.