Project definition
How Do We Understand the Relationship Between Two People?
A complete middle-layer method that separates character tendencies, interaction mechanisms, and situational storytelling.
Version: v0.1 | 2026-09-07 | Candidate perspective and method draft
Our starting point: character tendencies supply the raw material; the ways two people respond to each other constitute their relationship; situations and choices drive that relationship to change. One pairing name can hold many stories. Its greatest value is showing what problems these two people may encounter again and again when placed together.
This article presents this project's own framework. It draws on general relationship research in Psychological Foundations. The arrangement of the six specific mechanisms, the four-axis comparison, the naming system, and the approach to writing storylines were developed by this project and have not been validated as a relationship-prediction model.
1. What Each of the Three Layers Answers
| Layer | The reader's question | Responsibility | Status |
|---|---|---|---|
| Psychological foundations | What mechanisms in human interaction are worth understanding? | Explain the research questions, findings, methods, and scope | An accessible synthesis of existing research |
| Our middle-layer perspective | How can we use this knowledge to analyze these 16 characters? | Define our choices, observation steps, and creative method | A method proposed by this project |
| Pair-relationship layer | What is this pair called, and what can we discuss, write, or draw about them? | Provide a pairing name, two-way perspectives, scenes, and turning points | Character-based situational storytelling |
Readers can enter at any layer. Starting from a pairing name and moving downward reveals the relationship mechanisms it uses; going farther shows what the relevant research actually supports. Starting from the research and moving upward also shows where our own interpretation and creative work enter the process.
2. Four Questions, Asked of Each Person Separately
First expand A's and B's four axes. Each axis describes only a relative priority in relevant situations; it does not lock a person into the same response forever.
| Four-axis question | What to establish for A and B separately | What to ask when they are together |
|---|---|---|
| What do they contend for: Gain / Control | Do they prioritize securing material returns, or decision-making authority and confirmation of status? | Can both get what they need? Are they competing for the same thing? Who has the right to allocate it? |
| How do they push: Strategy / Pressure | Are they more likely to rearrange information, relationships, and options, or directly raise the cost of refusal? | Can the two approaches work in sequence? Do they undermine each other? Who bears the cost of acting openly? |
| When do they stop: Instrumental / Punitive | Once the external goal has been achieved, is there still value in making the other person pay? | Does one person think the matter is over while the other believes a debt remains? |
| How do they present it: Justification / Disregard | Do their actions or later explanations rely more on respectable, norm-based justification, or less on that kind of approval? | Who needs an explanation for outsiders? Who might support, exploit, or puncture that explanation? |
For example, Broker and Puppet Master are both Strategy, Instrumental, and Justification; their difference lies in Gain versus Control. This lets us pose a relationship question: if returns and decision-making authority can be satisfied separately, how do they trade? If Puppet Master changes the allocation to secure authority, how does Broker respond? That is a starting point for writing; the specific task and each side's leverage still need to be established.
Definition check. This draft follows *Core Model and Product Definitions v1.0*. Strategy may involve designing a situation openly, and Pressure may be applied through precise procedures. Some passages in the existing *Theoretical Foundations and Model Notes v0.1* frame Strategy as more covert; this draft does not adopt that narrower definition. Props and actions in the character assets may supply imagery, but concealed identities within those assets are not presented as real-world occupations or personality diagnoses.
3. Six Relationship Mechanisms
These mechanisms may occur simultaneously. Each relationship card selects the two or three that best explain its scene, helping readers explore further; the number of tags does not indicate relationship intensity.
M01 Collusion and the Contest for Position: What Does Each Person Want to Preserve in the Same Game?
Our perspective. To understand whether two characters can work side by side, first examine how their shared outcome is produced and who allocates it. The same goal may put people on the same path or make them compete for a single position; different goals may sometimes enable exchange. We place “whether both goals can be satisfied” and “who controls the conditions for achieving them” ahead of character similarity.
Four things deserve attention: whether a shared task truly exists; whether gains and decision-making authority can be separated; what each side controls; and how many alternatives either side has without the other. Actual power in a relationship arises from these conditions and cannot be read directly from the label “Control type.”
Original scenario. Broker provides access while Puppet Master coordinates the parties. They may need each other while the agreed gains and scope of authority are both respected. If that access becomes replaceable, the original allocation may be challenged. The reversal comes from a change in leverage; no one needs to switch type abruptly.
Writing prompts: “What can we accomplish together?” “Once it is done, who still needs to remain?”
Research pathways: interdependence theory and power-dependence analysis. S01, S02. The research supports analyzing interaction through situations and dependencies; the character storyline above is original to this draft.
M02 Mirroring and Complementarity: Recognizing Your Move—and Perhaps Following It
Our perspective. Similar behavioral languages readily create a sense of “you understand me,” but they can also lead two people to compete for the same position. Different methods may support a division of labor or produce a dispute over approach. We retain all four possibilities—shared language, shared blind spots, differences that enable coordination, and differences that cause friction—without assigning an overall positive or negative verdict to similarity or opposition.
Keep three things separate: how similar the two actually are; how similar they perceive each other to be; and whether their current behavior can work together. Dominance and affiliation in interpersonal complementarity research are also not equivalent to Control and Justification in our four axes.
Original scenario. Two Brokers both understand the unstated reservations in a deal, so negotiations can be economical. If each wants the other to reveal their true share first, that shared language turns into mutual wariness. Puppet Master and Enforcer may divide the work between arranging the route and driving execution, yet still dispute who is entitled to change the plan.
Writing prompt: “Can you do the part I cannot, or are you taking my place?”
Research pathways: goal and behavioral complementarity, interaction rhythms, and actual versus perceived similarity. S03, S04, S05, S06. These studies also suggest that coordination in interaction, subjective satisfaction, and long-term relationship quality should be discussed separately.
M03 Mismatched Ledgers: The Same Contribution Is Entered in Different Books
Our perspective. Many conflicts arise because the parties have not agreed on how the relationship should be accounted for. One treats help as an exchange, while the other sees it as care among one's own people; one talks about equal shares, while the other talks about seniority; one thinks payment settled the matter, while the other is still waiting for acknowledgment.
We examine the current rules through four plain questions: Is help given according to need? Is priority ranked by status? Is turn-taking or equal repayment expected? Are contributions and returns calculated proportionally? These questions draw on relational models theory, but real relationships may use different rules for different matters. Gain types cannot be reduced to exchange alone, and Control types cannot be treated as always demanding obedience.
Original scenario. Broker believes the agreed compensation has made good the loss. Smiler believes it repaid only the material account and did not address the slight. The same compensation may count as closure to one person and a brush-off to the other.
Writing prompt: “What do you think you have already repaid? What does the other person think you still owe?”
Research pathways: exchange and communal relationships; relational models theory. S07, S08. “Mismatched ledgers” is our accessible organizing term, not a separate psychological scale.
M04 Escalation and Ceasefire: How One Response Pushes the Next Round Further
Our perspective. Roles within a relationship change with the issue. First identify who is asking for change this time and who wants to preserve the status quo; then write the questioning, escalation, silence, delay, or counterattack. A Strategist may pursue intensely, and a Pressure type may refuse to respond. Do not map these two modes directly onto anxious or avoidant attachment.
Each cycle should specify at least three steps: what triggered it; how A interpreted and answered; and how B understood that answer. When writing a turn, add an observable stopping action, such as pausing further conditions, responding to one clear request, or returning the power to choose. A ceasefire, accepting consequences, and rebuilding trust may be separate stages.
Original scenario. A Arbiter presses for an answer about who will bear the loss, while the other person delays verification. The former reads delay as refusal to acknowledge the debt and adds demands; the latter reads escalation as proof that agreement is impossible and withdraws further. If they first separate confirmed items from disputed ones, the story may find an opening for a ceasefire. Simply falling silent cannot automatically be written as resolution.
Writing prompt: “If one person stops responding in the same way as the previous round, where does the story change?”
Research pathways: demand-withdraw patterns, related meta-analysis, and accommodation research. S09, S10, S11. This draft's “ceasefire conditions” are narrative and observational questions; they do not guarantee that any real relationship will be repaired.
M05 Frontstage and Backstage: The Audience Changes What the Same Words Do
Our perspective. An unseen third party may stand between two people: colleagues, friends, an organization, onlookers, or rules both recognize. Justification / Disregard describes different degrees of reliance on such approval; audiences, evidence, and the consequences of action also change how the two express themselves.
We ask: Who needs a defensible account? Who vouches for whom? Who will contradict the other in public? Can their private understanding withstand the consequences of becoming public? A shared line can coordinate action, but it may also conceal facts, so the actual conduct must be made clear as well.
Original scenario. Broker prepares an account of the allocation that every party can accept; Opportunist says outright who received more. This may be a comic act of puncturing the story, or the start of questioning an unequal allocation. If there is no conduct requiring justification, a person's wording alone is not enough to classify them as Disregard.
Writing prompt: “What do they know in private, and how much will each admit in public?”
Research pathways: impression management and moral disengagement. S12, S13. These concern presentation and self-restraint respectively; our method does not combine them into ready-made measurement evidence for Justification / Disregard.
M06 Exceptions and Mutual Shaping: This Time, Because It Is You, What Choice Did I Make?
Our perspective. What makes a character relationship memorable is often an exception with a concrete cost: choosing to state the real options clearly, allowing the other person to refuse, or supporting a goal the other person has chosen for themselves. An exception must be established through action in a scene; it cannot be calculated automatically from two type codes.
The two-way perspective matters especially here. A may already have changed a habitual response for B, while B reads it as ordinary cooperation. Both may be moving closer yet misreading how the other does so. Shared growth must identify who chose the goal of growth; “shaping you into what I prefer” cannot be written as the other person's self-actualization.
Original scenario. Puppet Master usually arranges every route in advance, but with one particular person presents the options and consequences together and accepts that person's choice of another route. This may begin an intimate bond, but the story must still show whether the other person chooses to respond and whether that room for choice remains later.
Writing prompt: “What would this person normally do? What changed this time? Did the other person recognize the change?”
Research pathways: relationship specificity, unique attraction, support for the ideal self, and shared novel activities. S14, S15, S16, S17. Shared experiences can broaden the material of a relationship; they cannot replace trust, responsibility, or addressing specific harm.
4. Six Additional Kinds of Situational Information
The four axes do not measure this information. Authors set it when writing fiction; analysis of real interaction requires observation or questions. If it is unknown, leave it unknown.
| Situational question | Minimum detail to establish | Why it changes the relationship |
|---|---|---|
| Shared task | What are the two trying to accomplish or obtain together? | The same pair may stand on different sides in a joint task and in a contest for one place. |
| Dependence and alternatives | Who controls information, resources, or permission, and who has other options? | Actual leverage may run counter to a character's preference for dominance. |
| Current issue | Who wants the status quo to change, and who currently benefits from it? | The pursuer and withdrawer can switch places. |
| Audience and rules | Who can see, and what account must hold up? | Private deals and public positions may constrain each other. |
| History and investment | What was owed, honored, or misunderstood in the past? | History can make the same behavior read as goodwill or repeated harm. |
| Autonomy and repair | Can either person refuse, pause, or leave; who will accept the consequences? | These choices determine how the story develops; type alone supplies no answer. |
5. Sixteen Same/Different Structures, Used Only to Find Questions
Recording whether the two match or differ in the order “goal, method, stopping point, norm” produces 16 structures. “Same” means only the same side of an axis, not the same strength, ability, position, or life experience. “Different” does not mean their behavior will oppose each other every time.
| Structure | Priority relationship question |
|---|---|
| Same–Same–Same–Same | With the same approach, who secures the position first, and can they recognize their shared blind spots? |
| Different–Same–Same–Same | Their methods are similar, but can material gain and decision-making authority be satisfied separately? |
| Same–Different–Same–Same | They want similar things but use different routes; can their actions connect? |
| Same–Same–Different–Same | At what point is the matter over, and does one side still think the price was insufficient? |
| Same–Same–Same–Different | Their actual goals and actions are similar, but do they have the same need for an outward explanation? |
| Different–Different–Same–Same | Their aims and methods differ; how can they find a shared task? |
| Different–Same–Different–Same | They can use similar methods, but do they disagree about the final prize and when to stop? |
| Different–Same–Same–Different | With different goals and public accounts, will they undermine each other within the same plan? |
| Same–Different–Different–Same | They want to reach the same place; how will they negotiate pace, route, and stopping conditions? |
| Same–Different–Same–Different | Their methods and need for respectability differ; who bears the cost of acting openly and explaining it? |
| Same–Same–Different–Different | They may proceed together at first; do they split over how to finish and explain it publicly? |
| Different–Different–Different–Same | Can a shared acceptable account still conceal their differences in goals and action? |
| Different–Different–Same–Different | Apart from a similar stopping point, what specific agreements do they need? |
| Different–Same–Different–Different | Does familiarity with the other's methods become a language for negotiation or a basis for anticipating each other? |
| Same–Different–Different–Different | Is their one similar goal enough to sustain the cooperation that follows? |
| Different–Different–Different–Different | With differences on all four axes, what external task or personal choice can bring them into the same game? |
For example, T01 × T09 differ only on the goal axis, so they are Different–Same–Same–Same. T13 × T16 match on goal and method but differ on stopping point and norm, so they are Same–Same–Different–Different.
The number of matching axes may be used as a completeness check. It cannot be added up into a compatibility, danger, or “good ship” score. Nor must a pair differing on only one axis have an easier relationship than one differing on all four.
6. How to Write a Relationship Card
- Expand the characters. Read both characters' four axes and role semantics; distinguish given tendencies from separately established abilities and experiences.
- Compare the four questions. Record commonalities and differences and select the two most worth writing about. Check the remaining axes as well so the copy does not contradict the definitions.
- Place them in a situation. Establish a task, a scarce resource, or an unresolved problem, and give the two different positions within it.
- Write each perspective. How might A understand B, and B understand A? Use “may” to mark these as candidate interpretations and avoid claims of mind-reading.
- Select mechanisms. Choose two or three. Their concrete operation should be visible in the structural rationale, three story branches, turning point, or ceasefire fields.
- Develop three branches. Show how cooperation becomes possible, how conflict emerges, and how an intimate exception is created through voluntary action. Each branch needs its own conditions.
- Find a turning point. Change one situational condition and observe how both respond anew. Then write one concrete ceasefire action instead of ending with “they communicate.”
- Condense it into a name. Choose an image from an object, action, space, or contradiction recurring between them. Let the opening line explain in one sentence why it belongs to this pair.
This sequence is a repeatable editorial check. The author must still give the task, leverage, and choices concrete form, then write them as observable interaction.
7. How Relationships Are Rewritten Over Time
We organize stories around four moments: entry, binding, fracture, rewriting. These terms locate story beats; a story may skip, reverse, or repeat any of them.
- Entry: Why do they approach each other—an opportunity, a shared task, a sense of similarity, or a need only the other can answer?
- Binding: Why do they stay? What does each gain, and what does each pay?
- Fracture: Which response is reinterpreted? Did the gains change, authority shift, an audience appear, or the original rules change unilaterally?
- Rewriting: What action moves the relationship toward cooperation, conflict, intimacy, distance, or an ending? Do both accept the new arrangement?
Take the “Leverage Pair” as an example: they enter through resource integration, bind through an exchange of gains and decision-making authority, and fracture when one side changes the allocation. Restating the terms and acknowledging the right to refuse may open another branch. Every arrow needs an action or condition as its bridge.
8. Naming for Recognition and Fan Creation
Pairing 16 distinct types produces 120 combinations; adding 16 same-type combinations yields 136. Each pair keeps one primary name, shared by A × B and B × A, while the two perspectives remain separate. The name covers the character combination. Partnership, rivalry, and close bonds are optional narrative branches beneath it. Branches on one card may represent different timelines; readers may develop one without combining every fragment into a single canonical story.
Names mainly use four approaches: a shared object, such as leverage or an outstanding balance; a continuing action, such as drawing the net closed; a shared space, such as adjacent seats or the same stage; or a central contradiction, such as a shared line punctured by one sentence. A name should sound natural, evoke the relationship, and accommodate multiple situations.
All pairing names in this version are project-created candidates. They are not presented as psychological classifications or disguised as community names already in popular use. If user-created aliases are collected later, record the name, source, author's attribution preference, and date collected separately. Adopting one as the primary name is an explicit content-version update.
The current primary checks for a name are: readers can identify the pair; one sentence can explain it; it differs from other names; and it still works when the storyline changes. A name cannot guarantee community uptake. Short chats with a few acquaintances serve only to find wording that is hard to read, misleading, or jarring.
9. A Note Website Readers Can See Directly
Every pair of characters has a name whose story can continue. Starting from what they contend for, how they move a situation, when they stop, and how they face rules, we organize their common ground, differences, and turning points. Partners, rivals, and close bonds are different directions their story may take; the specific relationship unfolds through the situation and both people's choices. To learn more, follow the mechanisms on a relationship card to read our analytical method and the related psychological research.
A short note beside each relationship card is enough: “A reference for creating character relationships; it does not predict real-world pairing outcomes.” Put the fuller discussion of research scope on the method and sources page so every passage need not repeat the same long disclaimer.
10. How Editors Judge Whether the Content Is Good Enough
Structurally correct. All 136 pairs are present, reciprocal lookup creates no duplicates, the 16 same-type pairs have clear situational positions, and axis meanings agree with the name mapping.
Concretely explained. After reading, someone can say what the pair does together, where a disagreement is most likely to emerge in the story, and at least one way a situational change could alter it. If removing the names leaves text that could be assigned indiscriminately to dozens of pairs, it still needs revision.
Perspectives come in pairs. A's goodwill cannot substitute for B's understanding, and A's desire for control cannot be treated as B's submission. A one-sided misreading can itself create a story, but the writing must identify whose misreading it is.
The turn is visible. Show change through actions such as returning options, stating conditions, or stopping additional demands. Do not use “they suddenly understood” or “it was destined” as explanations.
Sources are represented honestly. Research supports general mechanisms; the method and pairing content are identified as this project's proposals. Every research claim retains its original evidentiary status, and names do not serve as diagnoses or rankings of badness or relationship health.
Intimacy involves choice. A story of intimacy may contain conflict and darkness, but respect for refusal, acceptance of consequences, and voluntary support are available forms of relationship change. Dramatic tension alone does not show that real-world harm is worth enduring.
This round of source, content, and data checks is performed by Agent. If human feedback is used later, follow the project's established resource constraint: brief chats with no more than three or four acquaintances, focused on whether they understand the material and want to keep reading. Do not present this as statistical validation or psychometric validation.