X-Y Model of Dyadic Social Interaction

• X-Y model is a formal framework that defines dyadic relationships by categorizing interactions based on three actions (X, Y, ∅) selected by each agent.
• It systematically groups nine elementary action pairs into six distinct categories, including Equality Matching, Market Pricing, and Authority Ranking.
• The model aligns with Relational Models Theory by validating its core relational types and offering algorithmic classification of large-scale social data.

The X-Y model—also called the “X–Y model of dyadic social interaction”—formalizes the classification of pairwise social relationships by specifying all possible types of social exchanges between two agents, where each agent can select one of three actions in each interaction. Developed in direct dialogue with Relational Models Theory (RMT), this model provides a combinatorial and axiomatic basis for understanding the fundamental forms of human relational coordination. Its construction yields a rigorous taxonomy of possible dyadic relationships, categorizing them into six exhaustive and mutually exclusive types, four of which correspond precisely to the core relational models posited by RMT (Favre et al., 2013).

1. Definition of the X-Y Model and Action Set

Let $A$ and $B$ be two agents. Each agent $i\in\{A,B\}$ chooses a social action $a_i$ from: $\mathcal{A} = \{X, Y, \varnothing\}$ where $X$ and $Y$ are two distinct non-null actions (the choice is arbitrary but fixed, e.g., 'help', 'obey'), and $\varnothing$ represents 'do nothing.' An elementary dyadic event is represented as the ordered pair $(a_A, a_B)$, corresponding to the concurrent actions chosen by $A$ and $B$0 toward one another. Thus, the Cartesian product $B$1 yields 9 possible elementary configurations for each dyad: $B$2 These relationships are further organized into higher-level categories by logical grouping of configurations, as detailed below.

2. Exhaustive Taxonomy of Relationship Categories

The model systematically groups these nine dyadic action pairs into six exclusive categories, reflecting the possible qualitative types of relationships:

Category	Representative Pairs	Interpretation
1 (EM)	$B$3, $B$4	Equality Matching
2 (Null)	$B$5	Null interaction
3 (MP)	$B$6 and $B$7	Market Pricing (role-exchange)
4 (AR)	$B$8 or $B$9 (fixed order)	Authority Ranking (ranking)
5 (CS)	$i\in\{A,B\}$0, $i\in\{A,B\}$1 (or for $i\in\{A,B\}$2)	Communal Sharing (uncounted, reciprocal)
6 (Aso)	$i\in\{A,B\}$3 or $i\in\{A,B\}$4 (unidirectional)	Asocial (unilateral)

Where:

• EM = Equality Matching (bilateral, same action)
• Null = Mutual inaction
• MP = Market Pricing (exchangeable, action swapping)
• AR = Authority Ranking (unidirectional role, not swapped)
• CS = Communal Sharing (action reciprocated, not alternating)
• Aso = Asocial (single action, unreciprocated)

Each real-world dyadic interaction pattern, when decomposed, can be shown to correspond uniquely to one of these categories. The rigorous combinatorial structure excludes the existence of any further primitive (non-compound) types.

3. Generalization to Arbitrary Numbers of Actions

The construction remains valid and exhaustive when generalized to $i\in\{A,B\}$5 distinct non-null actions $i\in\{A,B\}$6 (plus $i\in\{A,B\}$7), with $i\in\{A,B\}$8. In this case, there are $i\in\{A,B\}$9 elementary dyadic pairs and $a_i$0 possible subsets. However, the same six categories persist:

• $a_i$1 diagonal cases correspond to expanded Equality Matching (EM).
• Unique $a_i$2 for Null.
• All unordered action pairs $a_i$3, $a_i$4, correspond to Market Pricing (MP) if both $a_i$5 and $a_i$6 are realized.
• Ordered off-diagonals correspond to Authority Ranking (AR).
• Exchangeable $a_i$7, $a_i$8 pairs for Communal Sharing (CS).
• Singleton $a_i$9 or $\mathcal{A} = \{X, Y, \varnothing\}$0 for Asocial.

This invariance under action set size implies that the structure of human dyadic relationships is constrained combinatorially irrespective of the complexity of the action set (Favre et al., 2013).

4. Correspondence with Relational Models Theory

Fiske's Relational Models Theory (RMT) posits exactly four coordinated models: Communal Sharing (CS), Authority Ranking (AR), Equality Matching (EM), and Market Pricing (MP), besides the degenerate asocial and null forms. The X-Y model demonstrates that these four, plus two, collectively form an exhaustive and necessary classification of dyadic interactions under minimal assumptions. Thus, the model provides a structural proof—rooted in combinatorics and symmetry arguments—of RMT's completeness for dyadic coordination (Favre et al., 2013).

5. Algorithmic Assignment and Data Application

Empirically, classification proceeds by algorithmic analysis of longitudinal dyadic data:

1. If both directions of distinct action (e.g., $\mathcal{A} = \{X, Y, \varnothing\}$1 and $\mathcal{A} = \{X, Y, \varnothing\}$2) are observed, classify as Market Pricing.
2. If only one directional alternation of distinct actions exists, assign Authority Ranking.
3. If repeated non-alternating single-action reciprocation, assign Communal Sharing.
4. If strict alternation of the same action, classify as Equality Matching.
5. If only unidirectional non-null action, classify as Asocial.
6. If only mutual inaction, classify as Null.

Tolerance for noise (occasional deviations) and time segmentation can refine classification, and coarse-graining can extend categorization to subgroups (Favre et al., 2013).

6. Value Matching and Relationship Stability

Agents may possess subjective value functions $\mathcal{A} = \{X, Y, \varnothing\}$3, determining their assessment of various exchanges. Stable coordinated relationships are hypothesized to satisfy 'value matching,' such as $\mathcal{A} = \{X, Y, \varnothing\}$4 for EM or CS. Persistent unequal valuations lead to relational instability. The model does not propose explicit utility functions but asserts systematic alignment between mutuality in action fluxes and equilibrium in perceived value exchanges (Favre et al., 2013).

7. Significance and Implications

The X-Y model establishes that the universe of dyadic social relationship types is finite, classifiable, and inherently constrained by the symmetry and combinatorics of individual action choices. It anchors RMT within a rigorous mathematical framework, answering the why of the “four relational models” and offering practical routes for the coding and analysis of large-scale social interaction datasets. Its sufficiency for classifying all forms of dyadic coordination underlies both sociological theory and the algorithmic processing of empirical social data. Any additional category must by necessity involve compound, mixed, or higher-order constructs not present in dyadic interactions with the minimal action set (Favre et al., 2013).