Chapter 17: Verbal Communication in Courtship — What We Say and What It Means

There is an old puzzle in the study of romantic attraction: people in the middle of falling for someone often report that they "just talked." They lost track of time. The conversation went on for hours. They can barely remember the specific topics. What they remember is a feeling — that the other person was there in a way that felt different from ordinary talking, that something happened in the language itself.

This chapter is about that something.

The science of verbal communication in courtship is, at its heart, about the gap between what language appears to do and what it actually does. On the surface, words transmit information: I tell you about my vacation, you respond with your opinion on Italian food, I ask whether you've ever been to Rome. But beneath that information exchange, language is doing something far more complex. It is negotiating closeness. It is building a shared world out of syntax and syllables. It is testing — carefully, often unconsciously — whether the person across from you is going to meet you where you are.

We will examine this process through multiple lenses in this chapter. We will look at the fascinating and counterintuitive research on linguistic style matching — the discovery that the function words we use (the ands, buts, pronouns, and prepositions) are better predictors of rapport than the content of what we say. We will investigate the psychology of self-disclosure and why the Aron "36 questions" study became one of the most-discussed social psychology experiments of the last thirty years. We will analyze the structure of compliments, the function of humor, the art of question-asking, and the revealing politics of who interrupts whom. And we will examine a phenomenon that may be the most distinctly human feature of courtship language: strategic ambiguity — the way that flirtatious speech works precisely because it can always be taken back.

Throughout, we will hold this material against two critical questions that run through the entire book. First: whose verbal styles count as charming, articulate, or appropriate in courtship contexts, and whose are pathologized or dismissed? And second: when verbal courtship becomes scripted — when "opening lines" are packaged and sold — what has happened to the communicative act we thought we were analyzing?

17.1 Language and Attraction: An Overview

The scientific study of verbal communication in courtship draws on at least four distinct research traditions, each with its own tools and preoccupations.

Sociolinguistics examines how social structure shapes language use — including differences by gender, race, class, and context. Foundational figures like Robin Lakoff (1975) and Deborah Tannen (1990) documented systematic differences in how men and women were socialized to use language, including in romantic contexts. Lakoff argued that "women's language" was characterized by hedges ("sort of," "I think"), tag questions ("...don't you think?"), and polite indirection — features that were often interpreted as weakness but which functioned as sophisticated politeness strategies. Tannen popularized the framework of "rapport talk" versus "report talk" — the idea that many women orient conversation toward intimacy-building while many men orient it toward information-exchange and status-positioning. These frameworks have been extensively critiqued and refined (more on this below), but they opened the door to taking gendered language seriously as an object of study.

Conversation analysis (CA), developed by sociologists Harold Garfinkel, Harvey Sacks, and Emanuel Schegloff, contributed the tools for examining the sequential structure of talk — how turns are allocated, how topics are opened and closed, how repairs happen when communication breaks down. CA-informed research on courtship has illuminated the micro-mechanics of interest signaling: the well-timed pause, the topic elaboration that signals "I want to keep talking about this," the question that opens rather than closes.

Communication science brought experimental methods: confederates trained to behave in standardized ways, coded interaction transcripts, surveys measuring perceived attraction after scripted exchanges. This tradition gave us most of the quantitative findings on compliments, question-asking, and humor in attraction.

Computational linguistics and text analysis, the newest entrant, gave us tools to analyze large corpora of naturally occurring courtship communication — dating app messages, speed-dating transcripts, couples' texting patterns. This is where Pennebaker's Linguistic Style Matching research lives, and it is arguably the most methodologically robust tradition because it can work with real communicative behavior rather than laboratory simulations.

What all four traditions agree on is that verbal communication in courtship is not merely information exchange. It is face negotiation, identity performance, relational work, and risk management — all simultaneously, often in the space of a single sentence.

17.2 Linguistic Style Matching: The Grammar of Rapport

Perhaps the most surprising finding in all of verbal communication research is this: if you want to predict whether two people are developing rapport, pay attention not to what they say but to how they say it — specifically, to the tiny grammatical glue words that hold sentences together and that virtually nobody pays conscious attention to.

This is the central insight of Linguistic Style Matching (LSM), a research program developed by psychologist James Pennebaker at the University of Texas at Austin and his collaborators.

17.2.1 What LSM Measures

LSM measures the degree to which two speakers synchronize their use of function words — a grammatical category that includes:

• Pronouns (I, you, we, they, it)
• Articles (a, an, the)
• Prepositions (in, on, with, about, through)
• Conjunctions (and, but, or, because, although)
• Auxiliary verbs (is, was, have, would, should, can)
• Negations (no, not, never, don't, won't)
• Quantifiers (some, all, few, many, more)

Notice what's not on this list: nouns, verbs, adjectives, adverbs. The content words — the parts of speech that carry most of the semantic information — don't appear in LSM calculations. This is deliberate. Content words are easy to consciously mirror: if you say "fascinating," I can repeat "fascinating" back to you without much effort. Function words are different. They reflect deep patterns of cognitive processing — how we locate things in time and space, how we relate ourselves to others, how we handle certainty and negation. They are largely automatic, processed below the threshold of conscious attention.

The LSM score between two speakers in a given conversation is calculated as the mean of seven component scores, one per function word category. For each category, the score quantifies how similar the two speakers' usage rates are. Perfect synchrony gives a score of 1.0; maximum mismatch approaches 0. In practice, strangers with no particular connection score around 0.65–0.70; couples in established relationships, and people who have just experienced a high-rapport interaction, often score above 0.85.

17.2.2 What the Research Shows

Ireland et al. (2011) conducted what remains the most direct test of LSM's role in courtship. They analyzed the instant-message conversations of over 80 heterosexual couples who had been recruited to participate in a speed-dating paradigm, then followed up. Their findings were striking:

• Higher LSM in the speed-dating message exchange predicted whether both partners reported mutual interest
• Among couples who went on at least one date, initial LSM predicted relationship length three months later
• The effect held even after controlling for the content of the messages — it really was the function word synchrony, not what people actually talked about, that did predictive work

In a separate study by Groom et al. (2011), small group interactions in laboratory settings showed that groups with higher average LSM had higher cohesion ratings and more productive problem-solving outcomes — consistent with the interpretation that LSM indexes genuine mutual engagement rather than mere politeness. What was particularly notable in Groom and colleagues' work was their examination of LSM across different communication modes: they found that LSM effects were present in both face-to-face and written communication contexts, though the absolute levels of synchrony differed. Written exchanges, which allow more time for reflection, showed slightly lower LSM scores on average — suggesting that the automaticity of function word use is somewhat attenuated when writing allows deliberate composition.

Research on couples' texting patterns has extended the LSM framework further. Studies examining text message exchanges in couples ranging from newly acquainted to long-term partners find that LSM scores track relationship development: early messages between two people who go on to form lasting relationships show higher LSM than early messages between people who do not progress. The effect is not large, but it is consistent and theoretically interpretable.

17.2.3 LSM in Different Contexts: Email, Text, and In-Person

The communication medium shapes LSM dynamics in interesting ways. In-person conversation, with its real-time processing demands, shows the most automatic synchrony — speakers have no opportunity to consciously monitor and adjust their function word use, so LSM reflects genuine underlying rapport most directly. Email introduces delays that allow reflection, but the LSM effect persists, perhaps because the underlying cognitive orientation toward the other person continues to shape language processing even with more time available.

Text messaging presents a particularly interesting case. Texting norms include a high degree of informality, abbreviation, and emoji use, all of which could potentially disrupt LSM measurement. Researchers who have adapted LSM scoring for text-message corpora find that the synchrony effect persists even when accounting for texting-specific conventions — suggesting that function word patterns survive the translation into the abbreviated register of text communication.

One dimension that emerges more clearly in digital contexts than in face-to-face research is the relationship between LSM and relationship longevity. Mairesse and colleagues found that LSM measured in early online communication was a significant predictor of whether online relationships survived beyond the first month of contact. The practical interpretation: linguistic synchrony in early digital courtship communication is not merely an epiphenomenon of attraction — it tracks something about the underlying compatibility of cognitive styles that predicts whether two people will sustain engagement.

17.2.4 Why Function Words?

Pennebaker's theoretical explanation is that function words are the grammar of relationship. When two people are truly engaged — listening carefully, thinking in terms of each other's perspective — they naturally begin processing language in more similar ways. The function word categories that shift most reliably with rapport include:

• Personal pronouns: People who are building intimacy together shift toward "we" framing; people who are emotionally distant maintain "I" and "you" as separate categories
• Negations: Interestingly, couples in conflict show high negation synchrony — they are "in sync" in their mutual denial and opposition, which is not the same as positive rapport
• Auxiliary verbs: Modal verbs (would, could, should) reflect shared orientation toward possibility and hypothetical futures — a space that often expands naturally in early romantic conversations

💡 Key Insight

LSM is not a technique that can be deployed strategically. The research consistently shows that attempted style matching — consciously mirroring someone's function word use — does not produce the same effect as the natural synchrony that emerges from genuine engagement. This is important: LSM is a symptom of mutual interest, not a cause. You cannot fake it reliably, which is precisely what makes it useful as a research measure.

17.3 The Python Analysis: Measuring LSM Yourself

The Python script accompanying this chapter (code/linguistic_style_matching.py) implements a working LSM calculator. It takes two text samples, tokenizes them into function word categories, calculates category-level usage rates, and produces an LSM component score for each category using the formula:

$$\text{LSM}_{\text{category}} = 1 - \frac{|r_A - r_B|}{r_A + r_B + \epsilon}$$

where $r_A$ and $r_B$ are the usage rates (proportion of total tokens) of the function word category for speakers A and B, and $\epsilon$ is a small constant to prevent division by zero.

The script compares two conversation pairs: a high-rapport exchange in which both speakers are clearly engaged and emotionally present, and a low-rapport exchange in which one speaker responds in an academic, distancing register. The bar chart output shows LSM component scores across all seven categories and an overall LSM comparison, making the difference visible.

📊 Research Spotlight: Running the Analysis

When you run the script on the provided conversations, the high-rapport pair typically scores 0.82–0.90 overall, while the low-rapport pair scores 0.65–0.72. The categories that diverge most sharply tend to be pronouns (the engaged speaker uses "I," "you," and "we" naturally; the distancing speaker uses more impersonal constructions) and auxiliary verbs (the engaged speaker uses modals freely; the academic register favors declarative constructions without modal hedging).

Try it yourself: collect a text exchange from a conversation you found particularly meaningful or particularly flat, paste the two sides into the script, and see what the numbers say.

17.4 Self-Disclosure and the Reciprocity Norm

If LSM tells us something surprising (that how we talk matters more than what we say), the research on self-disclosure tells us something more intuitive but equally important: the depth and symmetry of what we reveal to each other are among the most powerful determinants of whether attraction deepens into something more.

Self-disclosure refers to the sharing of personal information — feelings, experiences, beliefs, fears, desires — that goes beyond what is required by the immediate social situation. In courtship, self-disclosure serves multiple functions: it signals trust in the other person, creates opportunities for the other person to respond in kind (the reciprocity norm), and builds a shared fund of intimate knowledge that distinguishes the relationship from more casual acquaintance.

17.4.1 Jourard and the Foundational Research

Sidney Jourard's pioneering work in the 1960s established self-disclosure as a central concept in relationship science. Jourard (1964) argued, in terms that remain influential, that genuine self-disclosure — the willingness to make oneself known to another — is both a precondition for authentic relationships and a marker of psychological health. His early research documented substantial individual differences in disclosure willingness, and found that people disclose more to liked others and to others who have disclosed to them — establishing the reciprocity norm as a baseline empirical finding.

What Jourard observed was that self-disclosure is not simply information-sharing; it is, as he put it, an act of "letting oneself be known." The distinction matters: information can be transmitted without vulnerability; genuine self-disclosure involves the risk that what is revealed may not be received well. It is this vulnerability — not the informational content itself — that makes disclosure intimacy-generating.

17.4.2 Social Penetration Theory

Irwin Altman and Dalmas Taylor (1973) built on Jourard's work to develop Social Penetration Theory — one of the most influential models of how relationships develop over time. The theory uses the metaphor of an onion: relationships develop by peeling back successive layers of self-presentation, moving from the outermost layers (public, surface information) toward the core (deeply personal, highly private information).

Social penetration has two dimensions: breadth (the range of topics discussed) and depth (how personally revealing discussions of each topic are). Early in a relationship, breadth is typically greater than depth: two people talk about many different topics but at a relatively surface level. As intimacy develops, the pattern shifts — depth increases as conversations move from "what do you do?" to "what do you want out of your life?" to "what are you most afraid of?"

Altman and Taylor proposed that this progression is not unlimited. Social penetration encounters depenetration when the costs of continued intimacy — emotional risk, vulnerability, potential betrayal — outweigh the benefits. The theory predicts that relationships that develop too rapidly (skipping layers of depth) are often fragile, because the infrastructure of breadth and trust has not been built. In courtship contexts, this translates to a familiar phenomenon: the instant intense connection that burns out quickly because depth was achieved without the relational foundation to sustain it.

17.4.3 Digital Communication and Disclosure Dynamics

The digital transformation of courtship has meaningfully altered self-disclosure dynamics in ways that Social Penetration Theory, written in the pre-digital era, did not anticipate. Several changes are worth noting.

Asynchrony changes the risk calculus. In face-to-face interaction, self-disclosure requires real-time vulnerability — you cannot take it back, you see the reaction immediately, the social moment is live. In digital communication, asynchrony allows disclosure to be drafted, refined, and reconsidered before sending. This reduces the spontaneous vulnerability of disclosure while increasing its deliberate quality. The result may be more carefully calibrated disclosure, but less of the raw authenticity that genuine disclosure produces.

Anonymity and pseudonymity change disclosure norms. Research on computer-mediated communication has consistently found what researchers call the online disinhibition effect (Suler, 2004): people disclose more intimate information earlier in online relationships than in face-to-face ones, partly because reduced accountability (you cannot be immediately judged by someone who knows your social context) lowers the perceived cost of vulnerability. This accelerated disclosure can feel like rapid intimacy development — and it can be — but it can also produce what might be called parasocial premature depth: the sense of knowing someone well that outpaces the actual relational infrastructure of trust and shared experience.

Profile-based disclosure is a distinctive feature of dating app courtship. Before a conversation begins, each party has already disclosed a curated set of information: photographs, age, occupation, stated interests, perhaps a written bio. This pre-existing disclosure structure changes the disclosure dynamic of early conversation: the first message is already post-disclosure in some sense, which both compresses the early-stage breadth-establishment phase and creates pressure to demonstrate that one is consistent with one's profile presentation.

17.4.4 Aron's 36 Questions: Accelerating Intimacy

No study in this research area has captured public imagination as durably as Arthur Aron and colleagues' (1997) "Experimental Generation of Interpersonal Closeness." The study is popularly known as the "36 questions to fall in love" experiment, a characterization that is simultaneously not entirely wrong and significantly misleading.

📊 Research Spotlight: What Aron's Study Actually Found

Aron and colleagues (1997) developed a set of 36 questions, arranged in three "sets" of increasing intimacy. The questions range from relatively innocuous ("Given the choice of anyone in the world, whom would you want as a dinner guest?") to substantially more intimate ("When did you last cry in front of another person? By yourself?"). Pairs of strangers who worked through the questions — talking for roughly 45 minutes — reported significantly higher closeness on a self-report scale compared to pairs who engaged in small talk for the same duration.

The study became famous in 2015 when journalist Mandy Len Catron wrote a viral New York Times essay, "To Fall in Love with Anyone, Do This," describing her experience using the questions with an acquaintance she later married. The essay was enormously well-shared, and the questions are now widely circulated as a romantic "life hack."

What the original study actually shows is more modest and more interesting than the popular framing suggests:

1. Self-reported closeness, not love. The study measured "closeness" on the Inclusion of Other in Self scale — a reasonable but limited measure. Participants reported feeling closer; they were not tracked longitudinally for romantic outcomes.

2. The effect requires mutual participation. The questions work because of the reciprocity norm: each question invites genuine disclosure from both people. If one person treats the questions as an interrogation or responds with surface-level answers, the effect collapses.

3. Context matters enormously. The original study participants were unacquainted strangers in a controlled laboratory context who had consented to an "intimacy-generating" exercise. Using the questions on an unwilling or unsuspecting person is a different act entirely.

4. The questions aren't magic. What they do is structure a conversation so that disclosure naturally deepens — from medium to more personal to genuinely intimate. Any conversation that achieves that progression would likely produce similar results.

⚠️ Critical Caveat

The "36 questions to fall in love" framing has generated a small industry of articles, apps, and date-night cards that frame the questions as a technique for producing romantic feelings. This commodification is interesting in its own right (see Case Study 17.2 on pickup lines), but it also obscures what the research actually reveals: that genuine mutual disclosure is intimacy-generating, not the particular questions themselves. You cannot manufacture intimacy through a formula any more than you can fake LSM by counting conjunctions.

17.5 Compliments in Courtship

Compliments are among the most studied verbal acts in courtship research, in part because they are so structurally interesting. A compliment does at least three things simultaneously: it positively evaluates something about the recipient, it expresses something about the speaker's attention and preferences, and — crucially in courtship — it functions as an interest signal that creates an obligation to respond.

Research by Wolfson (1984) and Manes and Wolfson (1981), based on naturalistic observation of thousands of compliments in American English, found that compliments cluster around a remarkably small number of syntactic formulas. The most common: "That's a [adjective] [noun]" ("That's a beautiful scarf") and "[Pro-verb] [adjective] [noun]" ("You have a great sense of humor"). The content focuses overwhelmingly on appearance, ability/performance, and possessions — in roughly that order of frequency.

17.5.1 Gender Patterns in Complimenting

Several robust findings have emerged regarding gender and complimenting behavior:

• Women receive more compliments than men in most studied contexts, including heterosexual courtship settings
• Women give more compliments than men, particularly to other women
• The content of compliments differs: compliments to women focus disproportionately on appearance; compliments to men focus more on ability and possessions
• Men are more likely to interpret compliments as sexual interest signals than women are, a finding that has been replicated across multiple studies and attributed to differences in how the genders are socialized to parse ambiguous social signals

🔵 Ethical Lens

The appearance-focus of compliments directed at women reflects and reinforces a broader pattern in which women's social value is anchored in their looks in ways that men's is not. When a compliment on appearance is the default verbal move in heterosexual courtship, it participates in — even if it does not cause — the broader objectification dynamic discussed in Part VI of this book. This doesn't mean appearance compliments are inherently wrong; it means their prevalence is worth noticing and questioning.

17.5.2 The "Compliment Response Problem"

Wolfson's research also examined how people respond to compliments — and found that this is surprisingly difficult. In American English, the "correct" response to a compliment is neither straightforward acceptance (which risks sounding arrogant: "Yes, I do look great, don't I?") nor outright rejection (which insults the compliment-giver's judgment: "No, this shirt is terrible"). The preferred response, which Holmes (1988) called acceptance with mitigation, splits the difference: "Thanks, I got it on sale" or "Oh, I've had it forever." In courtship, this response pattern creates an interesting dynamic: the compliment has been acknowledged but not cashed in, leaving both parties somewhat suspended in the middle of a negotiation.

17.6 Humor and Rapport: An Introduction

Humor in courtship gets its own full chapter (Chapter 21), so here we introduce only the core framework. The key distinction in humor research is between shared laughter — genuine mutual amusement that arises from something both people find funny — and performed humor, in which one person deploys wit for effect.

Both matter in courtship, but they matter differently. Performed humor — being funny, making the other person laugh — correlates with attraction in early interactions, particularly for men (Lundy et al., 1998; Bressler et al., 2006). Being perceived as funny signals intelligence, social competence, and a willingness to take social risks. But shared laughter, the kind that arises organically from a mutually absurd situation or a callback to an earlier joke, is the better predictor of relationship satisfaction over time (Kurtz & Algoe, 2015). It is harder to manufacture and, like high LSM, tends to emerge rather than be deployed.

💡 Key Insight

The gender pattern in humor research is worth flagging: women report that "having a sense of humor" means the man makes her laugh, while men report that "having a sense of humor" means the woman laughs at their jokes. This asymmetry — documented by Bressler and Balshine (2006) — is a small, linguistically-embedded reflection of larger gender dynamics in heterosexual courtship scripts.

17.7 Question-Asking and Interest Signals

Questions are among the most powerful tools in early courtship conversation, and research by Karen Huang and colleagues (2017) at Harvard Business School produced a finding that, once stated, seems obvious but had not been empirically established: people who ask more questions are liked more.

Huang et al.'s analysis of speed-dating conversations found that asking questions — particularly follow-up questions that build on the previous response rather than switching topics — was one of the strongest behavioral predictors of whether the interaction partner reported wanting to go on a second date. Follow-up questions are significant for several reasons. They demonstrate that the asker was actually listening (not merely waiting for their turn). They signal genuine interest in the person rather than the topic. And they create a conversational structure in which the other person is positioned as someone worth knowing more about.

⚠️ Critical Caveat

Question-asking advice in popular media is often framed in crudely instrumental terms ("ask these 10 questions on a first date to get her to like you"). This framing misses the research finding: it is not the act of asking questions but the genuine curiosity behind follow-up questions that drives the effect. People can detect interrogation masquerading as interest; the research effect is about authentic attention, not formulaic question deployment.

Open versus closed questions also matter. Open questions ("What's that like for you?" "How did you end up in this field?") invite elaboration and personal reflection. Closed questions ("Do you like your job?" "Did you grow up here?") tend to close conversational threads rather than open them. In early courtship, the balance between open and closed questions both reflects and shapes the energy of the interaction.

17.8 The Grammar of Attraction: Syntactic Signals of Flirtatious Intent

Beyond the broad patterns of what people discuss in early courtship, linguists have begun to identify specific grammatical and syntactic markers that appear with elevated frequency in flirtatious conversation — features that signal romantic interest through the structure of language rather than its content. This line of research draws on computational approaches to classify flirtatious exchanges and identify their distinguishing features.

17.8.1 Second-Person Pronoun Frequency

One of the most consistent findings in this area is that flirtatious conversation shows elevated use of second-person pronouns — "you," "your," "yourself." The logic is relatively transparent: attraction creates an orientation toward the other person, and that orientation manifests in language that places the other person as the grammatical subject or object of sentences. Research by Cheng and colleagues on speed-dating transcripts found that both parties' use of "you" increased significantly in pairs where mutual interest was later reported, relative to pairs where only one party was interested or neither was. The you-orientation of flirtatious speech reflects, at the grammatical level, a cognitive orientation toward the other.

17.8.2 Hedging and Epistemic Modality

Flirtatious speech characteristically employs what linguists call hedging — language that softens assertions and marks them as provisional. Hedges include phrases like "I think," "sort of," "maybe," "kind of," and "it seems like." While hedging is sometimes interpreted as weakness or uncertainty, in courtship contexts it serves a face-protecting function: hedged statements are easier to retreat from if they are not received well. Saying "I kind of think you're interesting" rather than "I think you're interesting" creates a tiny margin of deniability.

This connects directly to the strategic ambiguity we examine below. Epistemic hedging — marking a statement as uncertain or provisional — is one of the primary grammatical mechanisms through which romantic interest can be signaled while maintaining plausible retreat.

17.8.3 Question Frequency and Conversational Asymmetry

Flirtatious conversation shows not only higher question frequency in absolute terms but a particular pattern of question asymmetry: one or both parties ask more questions than they receive, creating an interaction structure in which the question-asker is positioning the other person as the center of interest. This asymmetry is itself a signal — sustained, one-directional curiosity is not typical of neutral social interaction, and its presence communicates something about the questioner's orientation.

Research by Hall (2013) found that this questioning asymmetry was perceived differently by men and women: women rated sustained question-asking from men as more clearly signaling interest than men rated the same pattern from women, consistent with broader research showing that the default interpretation of social behavior differs by gender.

17.8.4 Temporal Markers and Future Orientation

Analysis of speed-dating transcripts shows that mutually interested pairs spend more time discussing hypothetical futures — conditional sentences, modal verbs oriented toward possibility, explicit references to future shared activities. "We should..." and "If you like..." and "Have you ever..." (inviting the other to envision experiences) create what researchers call shared future orientation, a grammatical construction of a possible "us." The appearance of this future-oriented language early in an interaction is a robust predictor of expressed mutual interest.

17.9 Topic Selection in Early Courtship

What do people actually talk about in early courtship, and what does topic selection reveal?

Research on speed-dating conversations (using transcription and coding) has found that topic complexity and novelty predict attraction better than topic commonality. Contrary to the folk wisdom that finding common interests drives connection, what seems to matter more is whether the conversation generates mutual discovery — topics where neither person has a fully formed answer, where they are thinking alongside each other rather than delivering rehearsed positions.

The topics that tend to generate mutual discovery in early courtship include:

• Hypothetical scenarios and counterfactuals ("If you could live anywhere for a year...") — these reveal values and imagination without requiring personal history disclosure
• Opinions with complexity ("I have complicated feelings about...") — demonstrating nuance signals intellectual depth
• Topics with some personal stake but not too much — the zone between trivial (favorite movies) and searingly intimate (childhood trauma) is where early courtship conversation tends to be most productive
• Future orientation — conversations about what each person wants, hopes for, and is curious about tend to generate more energy than retrospective accounts of past experiences

Topics that tend to flatten early courtship conversations include status-signaling content ("I went to a conference in Monaco last month"), complaints about absent third parties, and topics that require the listener to evaluate or judge people they don't know.

17.10 Strategic Ambiguity: The Art of Deniable Flirtation

Here we arrive at what may be the most intellectually distinctive feature of verbal courtship: the phenomenon that communication researchers call strategic ambiguity, and that most people experience as the strange, thrilling uncertainty of early flirtation.

Strategic ambiguity refers to the deliberate construction of verbal messages that can carry more than one interpretation — allowing the speaker to signal romantic interest while maintaining plausible deniability if the signal is not returned. The archetypal example is the comment that could be complimentary-but-harmless or flirtatious, depending on how the listener chooses to receive it. "You're always making me smile" is a nice thing to say to a friend. It is also something you might say to someone you are falling for. The ambiguity is not a bug; it is the entire feature.

17.10.1 The Logic of Deniability: Face Theory and Politeness

Why would anyone communicate with deliberate ambiguity when clarity seems more efficient? The answer is face protection. To understand this fully requires engaging with two foundational frameworks from sociology and sociolinguistics.

Erving Goffman (1967) introduced the concept of face to social theory: the positive social value that a person claims for themselves in interaction, and the ongoing collaborative work of protecting both one's own and others' face in social encounters. Goffman's insight was that face is always at stake in interaction, and that social actors are perpetually engaged in what he called face-work — the strategies, both linguistic and behavioral, through which face threats are managed, mitigated, and repaired. A direct, unambiguous declaration of romantic interest is a massive face-threatening act for both parties.

Building explicitly on Goffman, sociolinguists Penelope Brown and Stephen Levinson (1987) developed a comprehensive theory of politeness organized around the concept of face threat. They distinguished positive face (the desire to be approved of and liked) and negative face (the desire for autonomy, not to be imposed upon). Romantic overtures threaten both: the speaker's positive face is at risk if rejected; the listener's negative face is at risk because an explicit overture creates an obligation to respond.

Brown and Levinson's framework explains strategic ambiguity with precision: an ambiguous romantic signal threatens face far less than an explicit one, because its ambiguity allows both parties to interpret it in the least face-threatening way available. If the signal is not returned, it can be reinterpreted as friendly warmth — no face has been lost, no obligation has been created, no rejection has formally occurred. The interaction can continue as if nothing happened. This is not evasiveness; it is a sophisticated social technology for managing the inherently risky project of expressing romantic interest.

The practical operation of strategic ambiguity in conversation often involves extended touch, double entendres whose sexual interpretation can be plausibly denied, extended gaze that could be read as attentiveness or as something more, and phrases whose literal meaning is benign but whose pragmatic meaning — given tone, context, and relational history — carries romantic weight. The listener who wishes to acknowledge the signal can do so by responding in kind; the listener who does not wish to can simply respond to the literal meaning. Both parties understand the exchange; neither party is forced to make it explicit.

17.10.2 Decoding Ambiguous Signals

The challenge, of course, is that both parties are often operating under incomplete information about the other's interpretive frame. Research by Treat et al. (2012) and Abbey (1982) has documented systematic patterns in signal misreading — particularly the finding that heterosexual men, on average, tend to interpret friendly behavior from women as more sexually interested than women intend it. Abbey (1982) called this "the perpetual problem of the friendly female" — a framing that has been critiqued for attributing agency to the misreader rather than examining the structural conditions that produce the misreading.

The important point is that strategic ambiguity works as a communication system only when both parties share reasonably similar interpretive conventions. Cross-cultural, cross-gender, and cross-register differences in those conventions mean that what is intended as deniable flirtation can land as transparent aggression, or fail to register as anything at all. The system depends on shared code; without it, the very feature that makes ambiguity safe — its interpretability as either flirtatious or not — breaks down.

⚖️ Debate Point

Some feminist communication scholars argue that strategic ambiguity, while face-saving for the speaker, disproportionately burdens the listener — particularly women — who must decode signals without being able to ask for clarification without themselves risking face. The cost of the ambiguity, in other words, is not evenly distributed. Others argue that the face-protection function of strategic ambiguity is genuinely mutual and serves both parties. What do you think? Is strategic ambiguity a feature of polite social interaction, or does it primarily serve the interests of those with more social power to absorb the cost of an incorrect guess?

🔵 Ethical Lens: Ambiguity and Consent

Strategic ambiguity's relationship to consent is complicated and important. When verbal signals of interest are deliberately ambiguous, explicit consent-seeking becomes harder — not because people don't want to be asked, but because the communicative norms of flirtation were historically designed to avoid explicit asking. Consent researchers argue that this is precisely why cultivating direct communication norms matters: not because ambiguity is always wrong, but because ambiguity should coexist with the possibility of direct expression, not substitute for it entirely.

17.11 Interruption Patterns and Power Dynamics

Who interrupts whom, and what does it mean?

The classic study by Zimmerman and West (1975) found that in mixed-sex conversations, men interrupted women far more often than women interrupted men — by a ratio of roughly 96% of all interruptions coming from men in their sample. They interpreted this as a power dynamic: interruption, in their framework, is a dominance behavior that asserts the interrupter's right to the conversational floor over the speaker's right to complete their thought. Their sociological framing located this asymmetry within broader gender power structures, arguing that interruption patterns in conversation mirrored occupational and social hierarchies.

Subsequent research complicated this picture considerably. Linguist Deborah Tannen (1994) distinguished between intrusive interruptions (which silence the speaker and redirect the floor) and cooperative overlaps (which signal engagement and enthusiasm, not dominance — the verbal equivalent of leaning in). The problem with categorizing all interruptions as dominance behaviors, Tannen argued, is that it pathologizes high-involvement conversational styles that are culturally normative in some communities and that some speakers experience as pleasurable.

Research has confirmed that intrusive versus cooperative interruptions have very different social meanings and different distributions. In early courtship conversations, cooperative overlaps — finishing someone's sentence, jumping in to express agreement before they've finished — tend to be positively associated with mutual interest. Intrusive interruptions, which stop the other person mid-sentence to redirect, correlate negatively with reported interaction quality.

17.11.1 The Meta-Analytic Picture

Anderson and Leaper (1998) conducted a meta-analysis of 43 studies on interruption patterns in mixed-sex conversations and found the Zimmerman and West finding held for intrusive interruptions but not for all interruptions. Effect sizes were small to moderate (d ≈ 0.15–0.26), suggesting that gender accounts for some but far from all of the variance in interruption patterns. Task type, relationship type, conversational context, and individual personality all moderated the effect.

James and Drakich's (1993) landmark meta-analysis of research on gender and talkativeness — the most comprehensive review of the topic at the time — reached a similar conclusion: the folk belief that women talk more than men was not supported by the data; if anything, men talked more in mixed-sex public and task-oriented settings. The pattern of gender differences in interruption and floor-holding appeared to reflect not stable traits but contextual dynamics: who talks more (and who interrupts more) depends substantially on the social situation, the relative status of the interlocutors, and the conversational task.

In digital contexts, the interruption dynamic changes character entirely. Text-based asynchronous communication eliminates real-time interruption as a possibility — you cannot interrupt a text message. What digital courtship substitutes for interruption is response timing: how quickly you respond, how much you say, and whether your response engages with what was sent or redirects to your own concerns. Research on digital communication power dynamics is still developing, but preliminary work suggests that the person who consistently responds more briefly and with longer delays holds a version of the asymmetric communicative power that interruption captures in face-to-face interaction.

In courtship specifically, interruption patterns may be more symmetric, particularly in the initial stages when both parties are highly motivated to project interest. Some research suggests that mutual interruption in early courtship (cooperative overlapping) is actually a positive signal — two people racing to complete each other's thoughts, too excited to wait for their turn.

17.12 Cross-Cultural Verbal Courtship

Language is culturally embedded, and courtship language varies in ways that reflect deep cultural differences in the relationship between privacy and sociality, between directness and indirection, and between individual and group identity.

17.12.1 Direct vs. Indirect Communication Cultures

Linguist Edward Hall (1976) coined the terms high-context and low-context cultures to describe the degree to which communication relies on shared context versus explicit verbal statement. Low-context cultures (stereotypically: Germany, Scandinavia, much of the United States) tend to prize explicit, direct verbal communication — saying exactly what you mean. High-context cultures (stereotypically: Japan, South Korea, much of the Arab world, many parts of Latin America) embed more meaning in context, relationship history, and non-verbal cues, with verbal directness sometimes experienced as rude or even aggressive.

In courtship, this difference has substantial consequences. A study of verbal courtship patterns across cultures by Dion and Dion (1988) found that cultures with more collectivist orientations tended to have more indirect verbal courtship norms — where interest was signaled through sustained proximity, attentiveness, and contextual cues rather than explicit statements of interest. The same behavioral package — sustained attentive presence, interest in the other's activities, responsiveness to requests — reads as romantic interest in high-context cultures and potentially as friendship or courtesy in low-context ones.

In Japan, the linguistic concept of ma (間) — roughly, meaningful empty space — captures a cultural valuation of silence and restraint as communicative acts. Courtship language in Japanese contexts operates partly through what is not said: the refusal to be verbose, the elliptical statement that leaves interpretation to the listener, the gesture that communicates what direct speech would make embarrassingly explicit. A potential partner who "says too much" may be experienced as socially crude regardless of the warmth of what is expressed.

In contrast, Brazilian Portuguese courtship language has been documented as featuring high levels of explicit verbal expressiveness — direct compliments, verbal declarations of interest, and what might seem to Northern European interlocutors as inappropriately rapid emotional escalation. The same behaviors that signal warmth and genuine interest in São Paulo might signal desperation or social boundary-crossing in Stockholm.

17.12.2 Translation Failures as Cultural Revelation

Some of the most revealing evidence for the cultural specificity of verbal courtship norms comes from translation failures — the moments when a courtship utterance from one cultural context, translated into another language, loses its intended meaning or acquires unintended connotations. These failures expose the degree to which verbal courtship depends not just on linguistic content but on shared cultural scripts.

The German concept of Fingerspitzengefühl — roughly, "fingertip feeling," meaning exquisite sensitivity to situational nuance — has no direct English equivalent. In courtship contexts, German speakers sometimes describe the capacity for subtle, non-intrusive attention to a partner's signals as something that is performed through Fingerspitzengefühl, a concept that implies restraint, delicacy, and precise calibration. Translating this as "sensitivity" misses the specifically restrained, non-invasive quality of the German concept.

In Korean, the concept of nunchi (눈치) — the ability to read the room, to sense others' emotional states without being told — is a prized social quality that has direct implications for courtship. A person with high nunchi in a Korean courtship context can interpret indirect signals, pick up on unspoken discomfort, and calibrate their approach without requiring the other person to explicitly state their feelings. The communicative expectation is that a suitor should be able to read these signals; the need to ask explicitly can itself be read as social failure.

17.12.3 Verbal Modesty and Performed Indirection

Across many cultures, there exists a norm of verbal modesty in courtship that functions somewhat differently from strategic ambiguity. Where strategic ambiguity protects face by creating deniable signals, verbal modesty operates as a social script — an expected performance of uncertainty and reluctance that both parties understand is not literally meant. The conventional "I don't know, are you sure?" response to a romantic overture in some cultures is not genuine hesitation but a normative politeness move that signals virtue and respect for social propriety.

⚠️ Critical Caveat

The high-context/low-context distinction is a useful conceptual framework but risks sliding into cultural essentialism — the assumption that everyone from a "high-context" culture communicates similarly. In reality, there is enormous within-culture variation, individual differences are substantial, and the increasing globalization of communication norms (particularly among younger, urban, educated populations) means that any simple cultural generalization requires heavy qualification.

17.13 Verbal Self-Presentation: Storytelling in Courtship

One of the less-examined but most important verbal behaviors in early courtship is storytelling — the extended narrative sequences that allow one person to present a version of themselves through how they construct and tell a story.

Communication scholar Walter Fisher (1984) argued that humans are fundamentally "narrative animals" — that we understand ourselves and others through story rather than through argument. In courtship, this is acutely true. A first date is in large part an exchange of carefully curated personal narratives: how you ended up in your current city, the story of a formative experience, an anecdote that reveals character. These stories function not primarily as information transfer but as self-presentation — and skilled listeners evaluate not just the content of the stories but the way they are told.

Skilled courtship storytelling tends to share certain features: a clear narrative arc, a moment of genuine vulnerability or surprise, some degree of self-awareness or humor about one's own role in the story, and an ending that opens toward the listener (a question, a reflection, an invitation to respond). Stories that are purely self-aggrandizing — where the teller always wins and is always right — tend to land poorly, even when the specific content would seem impressive.

💡 Key Insight

Research on narrative self-disclosure (Reis & Shaver, 1988) suggests that the most intimacy-generating stories in early courtship are not the most impressive ones but the ones that reveal something genuine — a failure, a confusion, an ongoing question that hasn't been resolved. This aligns with the broader finding that vulnerability is the engine of intimacy: not the performed vulnerability of calculated confession, but the authentic kind that involves not knowing how the listener will respond.

17.14 The Evolution of Courtship Language: From Letters to Texts

The history of courtship language is also a history of the media that carried it — and the constraints those media imposed.

The formal letter was the dominant medium of long-distance courtship from the early modern period through most of the 20th century. Letters imposed their own discipline: the writer had time to revise, the conventions of formal address created distance and formality, and the delay between sending and receiving meant that courtship proceeded in deliberate, measured steps. Linguists who study historical love letters note both the formality (endearments were carefully structured within conventional forms) and the extraordinary verbal elaboration — long letters were evidence of effort and investment.

The telephone introduced immediacy and voice but retained the mediated quality of distance. The answering machine introduced asynchrony back into immediate-medium communication — you could leave a message and manage the timing of your words.

Digital communication — email, then text messaging, then direct messages across multiple platforms — introduced a new kind of asymmetry: messages that arrive immediately and invisibly await response, creating both the intimacy of presence and the anxiety of uncertainty. The read receipt may be the single most psychologically significant feature of digital courtship communication: the moment when you can see that your message has been read but the response has not yet come.

Texting has developed its own paralinguistic conventions — ellipses that convey hesitation, capitalization for emphasis, the use of lowercase as aesthetic casualness, the precise management of emoji use to signal warmth without commitment. Linguistic researchers note that texting has not degraded courtship language but created a new vernacular with its own rich set of courtship-specific conventions (Thurlow & Mroczek, 2011).

🔗 Connection to Other Chapters

Chapter 20 examines dating app communication in depth, including the opening message problem (how do you start a conversation with a stranger?) and research on message response rates. The historical arc traced here — from letters to texts — provides essential context for understanding why the communicative norms of digital dating platforms feel simultaneously new and familiar.

17.15 Intersectionality and Verbal Courtship Norms

Before closing, we need to interrogate a feature of this entire chapter that should be making you uncomfortable: most of the research reviewed here has been conducted on relatively homogeneous populations, and the norms it identifies as "how people communicate in courtship" are often, in practice, the norms of white, educated, Western, heterosexual courtship.

When sociolinguistics talks about "gender differences in communication," whose gender is being described? The binary framework itself excludes nonbinary and trans individuals, and the behavioral patterns attributed to men and women are often more accurately described as patterns associated with masculinity and femininity as cultural constructs — which means they vary by race, class, and cultural context in ways that the gender-binary framing obscures.

When communication research talks about "directness" and "appropriate self-disclosure" in early courtship, it often encodes middle-class professional communication norms as universal. Research on code-switching — the practice of shifting between linguistic registers depending on social context — shows that many people of color, particularly Black Americans, navigate fundamentally different communication environments than their white counterparts, including in romantic contexts (Baugh, 2000). The verbal style that reads as "confident and charming" in one context may read as aggressive or inappropriate in another, depending on racialized perception norms that have nothing to do with the speaker's actual communicative intent.

⚖️ Debate Point

Deborah Tannen's popular "men talk for information, women talk for connection" framework has been widely influential and is based on genuine research — but critics including scholars of race and class argue that it over-generalizes patterns specific to white, middle-class American culture. Toni Morrison wrote about the entirely different norms of connection-building in African American vernacular traditions. Whose verbal styles count as "relationally sophisticated," and whose count as "abrasive" or "too much"? These evaluations happen constantly in courtship contexts, with real consequences for who is deemed attractive and who is not.

17.16 The Commodification of Courtship Language

One of the recurring themes in the study of verbal courtship is the persistent attempt to systematize, package, and sell what are fundamentally spontaneous, relational communicative behaviors. From Victorian-era etiquette guides to twentieth-century pickup artistry manuals to the AI-generated icebreaker suggestions now built into dating apps, there has been a persistent market for scripts, formulas, and techniques that promise to reduce the uncertainty of verbal courtship.

The appeal is obvious: verbal courtship is uncertain, face-threatening, and dependent on skills that are unevenly distributed. The promise of a reliable opening line, a sequence of conversation moves that produces attraction, or a formula for self-disclosure that generates intimacy — these promises reduce the anxiety of genuine communicative risk. And the market for such promises is enormous.

What does the research say about scripted courtship language? The findings converge with the broader pattern identified throughout this chapter: scripted, formulaic verbal approaches consistently underperform authentic, contextually responsive communication. Studies on opening messages in dating contexts find that generic openers — the ones most likely to come from a formula or script — have dramatically lower response rates than messages demonstrating genuine attention to the specific person and profile. The value of the personalized, contextually specific message is not its informational content but what it signals: that the sender actually read, and actually noticed, something specific to this particular person.

The algorithmic icebreaker represents the contemporary endpoint of this commodification. Dating apps now offer auto-generated opening lines, with language models analyzing profile content to suggest personalized-seeming openers. This creates a striking paradox: the message that appears to signal genuine attention may be the output of a statistical model, not the expression of authentic curiosity. The appearance of engagement is severed from actual engagement, and the communicative signals that generate rapport — LSM, genuine self-disclosure, follow-up questions that demonstrate actual listening — become harder to interpret when they can no longer be assumed to reflect genuine underlying interest.

🔵 Ethical Lens: The commodification of courtship language is, at its limit, a project of manufacturing the appearance of the qualities that make verbal courtship genuinely intimacy-generating. The qualities most reliably associated with attractive verbal behavior — authentic attention, genuine curiosity, willingness to be vulnerable — are precisely the qualities that cannot be scripted without hollowing out what makes them effective. The industry that promises to script them sells the simulation of the thing rather than the thing itself.

Chapter Summary

This chapter has examined verbal communication in courtship across multiple dimensions. We began with the counterintuitive discovery of Linguistic Style Matching — that unconscious synchrony in function word use is a better predictor of rapport than any content dimension, with effects extending across email, text, and in-person contexts. We examined Jourard's foundational self-disclosure research and Social Penetration Theory's layered model of intimacy development, noting how digital communication has altered the timing and texture of disclosure dynamics. We analyzed Aron's 36 questions study, noting what the research actually shows versus what its popular framing suggests. The grammar of attraction section identified specific syntactic markers — second-person pronoun frequency, hedging, question asymmetry, future orientation — that distinguish flirtatious from neutral conversation. We examined compliments, humor, question-asking, and topic selection, finding in each case that the behaviors associated with successful courtship are those that express genuine attention and interest rather than formulaic execution.

We examined strategic ambiguity — the way that flirtatious speech maintains plausible deniability, grounded in Goffman's face theory and Brown and Levinson's politeness framework — and found it to be a genuinely sophisticated communicative phenomenon with real consequences for consent and face protection. We looked at interruption patterns (including James and Drakich's meta-analytic contribution), storytelling as self-presentation, cross-cultural variation in verbal courtship norms (including ma, nunchi, and the translation failures that reveal cultural specificity), and the historical evolution of courtship language across media. Finally, the commodification section examined how the packaging and sale of courtship language formulas has altered the communicative landscape in ways that complicate the interpretation of even genuine attention signals.

Throughout, three themes from the book's overall arc ran through the analysis: the question of agency and consent in verbal signaling, the intersectional critique of whose verbal styles count as appropriate, and the commodification of courtship language — from scripted pickup lines to algorithmic message suggestions — that turns the intimate act of verbal connection into a product.

What language does in courtship is ultimately not reducible to information transfer. It is the medium through which two people find out whether a shared world is possible. And that finding-out, the research suggests, happens not primarily in what we say but in the texture of how we say it together.

Key Terms

Linguistic Style Matching (LSM) — The degree to which two speakers synchronize their use of function words; a measure of verbal rapport developed by Pennebaker and colleagues.

Function words — Grammatical words (articles, pronouns, prepositions, conjunctions, etc.) that structure sentences; contrasted with content words that carry semantic meaning.

Self-disclosure — The sharing of personal information beyond what a situation requires; central to intimacy development.

Social Penetration Theory — Altman and Taylor's (1973) model of relationship development through progressive self-disclosure, using the metaphor of an onion with breadth and depth dimensions.

Reciprocity norm — The social expectation that disclosure will be met with equivalent disclosure, creating a mutual deepening of intimacy.

Online disinhibition effect — The tendency for people to disclose more intimate information earlier in online relationships than in face-to-face ones, due to reduced accountability and anonymity.

Strategic ambiguity — The deliberate construction of verbal messages with multiple possible interpretations, allowing interest to be signaled while maintaining deniability.

Face — In Goffman's and Brown and Levinson's frameworks, the public self-image that social actors need to maintain; threatened by explicit rejection. Distinguished as positive face (desire to be liked) and negative face (desire for autonomy).

Epistemic hedging — Language that marks statements as provisional or uncertain ("I think," "sort of," "maybe"); a grammatical mechanism that creates deniability in flirtatious communication.

High-context vs. low-context cultures — Hall's distinction between cultures that embed meaning in context (high-context) versus those that rely on explicit verbal statement (low-context).

Cooperative overlap — An interruption that signals engagement and enthusiasm rather than dominance; distinguished from intrusive interruptions that silence the speaker.

The Python script for this chapter (code/linguistic_style_matching.py) implements a working LSM calculator. See the exercises for hands-on applications.