1. Motivation

Despite decades of progress in digital publishing, the scholarly article remains a fundamentally impoverished medium (1, 2). While JATS XML elegantly captures structure, metadata, and cross-references, it entirely ignores the olfactory dimension of knowledge. As illustrated in the Key Image, the ancient process of olive pressing — crushing, filtering, refining — offers a compelling metaphor for the transformation of raw scholarly content into structured, multi-sensory digital objects. This is a striking omission given that smell is the sense most strongly linked to memory formation and emotional engagement (3) — precisely the qualities that scholarly articles most lack.

2. OOML Specification

OOML extends JATS 1.4 with a dedicated namespace (xmlns:ooml) and 14 new elements (4), including <nose>, <palate>, <finish>, <harmony>, <defect>, <pairing>, and <drizzle> (the full element set is specified in the Appendix). The layered encoding architecture is illustrated in Figure 1. The core innovation is the <aroma-profile> element, which encodes scent descriptors from the International Olive Council’s standardized vocabulary (5): fruity, bitter, pungent, with optional sub-notes of artichoke, green banana, freshly cut grass, and tomato leaf. Each section of an article can carry its own <tasting-note>, allowing the reader’s olfactory experience to evolve with the argument.

3. User Study

Three hundred forty researchers across 12 disciplines evaluated OOML-enhanced articles delivered via scratch-and-sniff e-ink displays. The scratch-and-sniff e-ink display prototype (model OliveInk™ v0.3) was manufactured by the Kalamata Institute of Advanced Peripherals. Only 12 units were produced; 3 were consumed during calibration. The study followed a three-phase protocol:

1. Participants received randomized article pairs (OOML-enhanced vs. unscented control) in a double-blind configuration, where the control group’s aroma diffuser dispensed distilled water.

2. Reading sessions lasted 45 minutes with standardized Arbequina diffusion at 22°C and 40% relative humidity — conditions established by Pressington (1) as optimal for scholarly olfaction.

3. Engagement was measured via eye-tracking, page-turn latency, and a self-reported “Would You Cite This?” score on a 7-point Likert scale.

Reader engagement was modeled using the Olfactory Engagement Function (6):

where p is the pungency index (0–10) and λ is the reader’s olfactory sensitivity coefficient, empirically determined to average 0.42 across disciplines (7). At the optimal pungency of 7.2, this yields a completion probability of 95.3% — compared to 34% for unscented control articles. Full results by discipline are reported in Table 1.

As predicted by Equation 1, the effect was strongest in mathematics, where participants reported that a hint of Arbequina olive oil made proofs feel “more intuitive” (8). One participant offered the following testimony:

“I had been stuck on a proof of the Riemann Hypothesis for eleven years. After reading the OOML-enhanced version of my colleague’s paper with a gentle Koroneiki diffusion, the solution came to me in minutes. Unfortunately, the proof evaporated along with the volatile compounds before I could write it down.”

4. Implementation

OOML is fully backwards-compatible: standard JATS parsers ignore the ooml: namespace, while OOML-aware readers can render the sensory layer (9, 10). We provide an XSLT stylesheet that compiles OOML to both standard HTML5 (visual only) and to our experimental OliveReader format, which drives USB-connected aroma diffusers. A prototype implementation is being tested on Olive Turn Scholar Core (oliveturn.com), where the platform’s submission bundle accessibility auditing tools have provided early data on the challenges of multi-sensory content validation.

The transformation pipeline processes each <tasting-note> element into a timed aroma instruction sequence. For example, the following OOML fragment:

compiles to an OliveReader instruction that activates scent cartridge channels 3 (green/herbal) and 7 (peppery) in sequence, with channel 7 sustained for 45 seconds at reducing intensity (11). The polyphenol concentration threshold for reliable olfactory activation was determined to be 10³ mg·kg⁻¹.

5. Accessibility Considerations

JATS 1.4 continues to uphold accessibility support through elements such as <alt-text> and <long-desc>, ensuring that visual content carries textual descriptions for assistive technologies. Aligning these with WCAG 2.2 AA provides a comprehensive framework for visual and auditory accessibility. However, both standards remain silent on olfactory and gustatory content — a gap that OOML exposes with some urgency. The current state of multi-sensory standards coverage is summarized in Table 2.

We propose <aroma-alt-text>: a short textual description of encoded scents for readers who cannot smell, analogous to <alt-text> for images (12). For example: <aroma-alt-text>Warm, grassy notes of fresh Koroneiki oil with a peppery finish</aroma-alt-text>. For richer descriptions, we propose <gustatory-long-desc>, providing extended taste narratives compatible with screen readers and refreshable Braille displays.

Early testing with Olive Turn Scholar Core’s submission bundle accessibility auditing tools flagged 100% of OOML-enhanced articles as containing “unknown sensory content with no textual equivalent” — confirming that current tooling is unprepared for multi-sensory scholarship. We urge the W3C and NISO to consider extending their accessibility guidelines to accommodate the olfactory and gustatory dimensions of knowledge before the first scratch-and-sniff e-ink displays reach general availability.

6. Understanding

We wish you a happy first of April and confirm — if not yet apparent — that this article is an April Fools’ Day contribution. While OOML remains fictional, the challenges of accessibility in scholarly publishing are very real. We hope this lighthearted exploration draws a smile and perhaps a moment of reflection on how far markup languages still have to go.