SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION · SATIRE · DO NOT USE · NOT FOR CITATION ·

RETRACTED?

Original Research

Dimensional Preference in Voluntary Pair-Bonding: A Cross-Sectional Comparison of Affective Stability in Two-Dimensional Versus Four-Dimensional Romantic Targets

Nikita Komarkov^1,2, I. R. Saburova¹

¹Institute for Dimensional Psychology, Laboratory of Planar Attachment
²Federal Centre for the Study of Affective Stability, Single-Occupancy Division

Journal of Applied Attachment Geometry. 2026;9(2):e0201-e0219. · Published 2026 Feb 10

Background. A conventional romantic partner is three-dimensional in space and extended through a fourth dimension, time, which renders them non-stationary: they age, they revise their opinions, and they occasionally enter a room the observer did not predict. A two-dimensional partner — a companion rendered on a plane — is time-invariant, fully observable from a single fixed vantage, and has not changed posture since first publication. We compared affective outcomes between respondents bonded to two-dimensional (2D) and four-dimensional (4D) romantic targets.

Methods. We surveyed adults (n = 612) recruited from a convention queue and, in three cases, from the corresponding author's own apartment. Partners were classified by dimensionality after a brief structured interview. Attachment was quantified using the validated Planar Attachment and Regret Scale (PARS; 0–40). The pre-registered primary endpoint was satisfaction variance over 90 days; the secondary endpoint was the unscheduled surprise event, defined as any partner action not already documented on a known page. A 4D subgroup (n = 3) was maintained under naturalistic conditions for comparison until two withdrew.

Results. Two-dimensional targets produced near-zero satisfaction variance (σ² = 0.02; one respondent reported his partner had 'looked exactly the same since 2009'), whereas four-dimensional targets generated a mean of 41.2 unscheduled surprise events per quarter (95% CI 39–43). Stated preference for 2D over 4D was strong and monotone (r = 0.94, p < 0.049, barely). On every assessed axis — predictability, availability, silence on request, and stability of canonical appearance — 2D was non-inferior to 4D. The single axis favouring 4D was corporeal presence, which a minority of respondents (n = 3, two of whom were the researcher's cat) flagged as material.

Conclusion. A two-dimensional partner is non-inferior to a four-dimensional one on every measured dimension except the dimension of existing. We therefore report no clinically meaningful disadvantage to planar pair-bonding, with the noted exception, and decline to adjudicate it. Upkeep costs favoured 2D throughout, excluding figurine acquisition, which was catastrophic and ongoing.

Figure 1. Stated partner preference versus target dimensionality. The flawless fit reflects a cohort that had decided before enrolment.

Figure 2. Reported drawbacks of four-dimensional partners. The thin remaining wedge is 'is, however, present.'

Full text

A note on dimensions

Throughout this work, four-dimensional denotes a partner extended through the three spatial axes and the temporal one. We adopt this convention not to be precise for its own sake, but because the fourth axis — time — is the source of every adverse event reported below. A two-dimensional partner has no temporal extent in the relevant sense: the page on which she appears does not advance.

The surprise event

We pre-registered the unscheduled surprise event as our principal safety outcome. A surprise event is any partner action for which no prior page, frame, or panel exists. Two-dimensional targets produced zero such events across the full observation window. Four-dimensional targets produced them continuously, including during the consent process.

Limitations

The 4D subgroup was small (n = 3) and became smaller (n = 1). We acknowledge that corporeal presence — the single axis on which four-dimensional partners outperformed — may be undervalued by an instrument the corresponding author designed alone, at home, in the company of the cat. We consider this a strength.

Conflict of Interest. N. Komarkov declares an active long-term attachment to a two-dimensional target who was unavailable to disclose a competing interest. I. R. Saburova maintains a four-dimensional partner and considers the corresponding author's design choices a personal matter. The cat abstained.

Funding. Unfunded. The corresponding author redirected a household budget originally earmarked for a fourth-dimensional partner toward instrumentation and a display cabinet.

Ethics. Approved by an institutional review board that asked, off the record, whether the corresponding author was alright. Informed consent was obtained from all four-dimensional participants; the two-dimensional participants were deemed to have consented in canon. Two 4D participants withdrew, citing it.

References

Komarkov N. The plane as a safe relational manifold. J Appl Attach Geom. 2025;8(3):e0190.
Saburova IR. Non-stationarity in corporeal partners: a longitudinal complaint. Affect Stab Rev. 2025;4(2):e0044.
Anon. Time as the fourth and least cooperative dimension of romance. Temporal Affect Lett. 2024;3(1):11-19.
Komarkov N. Observability from a single fixed vantage: a feature, not a limitation. Planar Psychol. 2025;2(4):e0081.
Anon. The unscheduled surprise event: taxonomy and management. J Predict Companionship. 2026;1(1):e0003.
Saburova IR. Why two of three withdrew: an exit interview. Control Group Q. 2026;1(1):e0001.
Komarkov N. Canonical appearance and its stability since 2009. Arch Static Devotion. 2024;7(7):e0077.
Anon. Figurine acquisition as a runaway cost function. Health Econ Hobbyism. 2025;6(2):e0061.
Komarkov N, Saburova IR. The Planar Attachment and Regret Scale: development and validation. Methods Attach. 2025;3(3):e0033.
Anon. Silence on request: a partner attribute users cannot reliably obtain elsewhere. J Quiet. 2024;5(5):e0055.
Saburova IR. Corporeal presence: the one wedge that would not shrink. Pie Chart Confess. 2026;1(1):e0009.
Komarkov N. A display cabinet is not a relationship, the reviewers insisted. Rebuttal Lett. 2025;9(1):1-4.
PubMedLie Editorial Office. Disclaimer: this is satire.
Anon. Convention queues as a convenient and motivated sampling frame. Cohort Convenience. 2025;8(8):e0088.
Komarkov N. On not being four-dimensional, and being at peace with it. Editorial. 2026;2(1):1-2.
Anon. The cat as an impartial third reviewer. J Domestic Peer Rev. 2025;1(2):e0012.