📄 social_science

Jealousy Without a Face: A Systematic Review and Meta-Analysis of Nonverbal Signals of Loss and Threat

This preregistered systematic review and meta-analysis of ten effects from nine studies (N ≈ 1,847) reveals that while jealousy lacks a distinct nonverbal signature, it is expressed through a candidate architecture of anger, sadness, and distress signals, highlighting a significant gap in direct research on jealousy's involuntary manifestations.

Original authors: Maria Barbato, Ana María Fernández, Michele Dufey, Belen Zavalla, Teresita Campino

Published 2026-08-03

📖 8 min read🧠 Deep dive

CC BY 4.0

Original authors: Maria Barbato, Ana María Fernández, Michele Dufey, Belen Zavalla, Teresita Campino

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your brain is a super-smart detective that never sleeps. It's constantly scanning the world for clues about what people are feeling, not just by listening to their words, but by reading their body language. This field of science is called nonverbal communication. Think of it like a secret code written in facial expressions, eye movements, and how someone holds their shoulders. Some clues are like a frozen photograph—a quick snapshot of a face that might look angry or sad. Others are like a movie clip, showing how a person's head tilts, how they shift their weight, or how their eyes dart around over time. Scientists have long known that emotions like anger and sadness have their own unique "signatures" in this code. But there's one emotion that is incredibly powerful in our social lives, yet remains a bit of a mystery in this visual language: jealousy. We all know what it feels like when we're worried about losing someone we care about to a rival, but we don't really know what jealousy looks like on a face or in a body. Is there a specific "jealous face"? Or is it a messy mix of other feelings?

This paper, titled "Jealousy Without a Face," is a massive detective story where the researchers didn't just look at one case; they gathered evidence from nine different studies involving nearly 1,850 adults to solve the puzzle. Instead of looking for a single "jealousy face" that doesn't seem to exist, they investigated the building blocks of jealousy: anger, sadness, and distress. They wanted to see how these feelings show up in our bodies, both in those frozen snapshots (static signals) and in the moving movie clips (dynamic signals). The big question was: if jealousy is a mix of these feelings, can we spot the ingredients? The researchers found that there isn't one single, magical "jealousy expression" waiting to be discovered. Instead, they found a "loss-plus-threat" architecture. It's like a recipe where the main ingredient is sadness (the feeling of loss), which shows up clearly in both photos and videos as a downward gaze and a slumped posture. But layered on top of that sadness are "threat" signals, mostly visible in static photos as a redder face or a sharper, more angular jawline, which signal anger. The study suggests that jealousy is best understood as this complex, layered combination of feeling hurt and feeling threatened, rather than a simple, standalone emotion with its own unique face.

The Detective Work: How They Solved the Mystery

To understand how the researchers cracked this case, imagine they were trying to find a specific flavor in a giant bowl of soup. They knew the soup was supposed to taste like "jealousy," but they couldn't find a single spoonful that tasted exactly like that. So, they decided to taste the main ingredients separately: anger, sadness, and distress. They gathered data from nine different scientific studies (with a total of about 1,847 people) that had already tested how people express these specific emotions without using words.

The researchers used a very strict set of rules to pick their studies, making sure they only looked at experiments where they could actually see or measure the body language, not just ask people "How do you feel?" They looked at two types of clues:

Static Signals: These are like frozen photos. Think of a picture of a face that looks red, or a face with a very sharp, angular jaw, or a specific set of wrinkles around the eyes.
Dynamic Signals: These are like video clips. Think of someone looking down, shrinking their body, or making tiny, quick movements that show they are being honest or trying to hide something.

The Big Findings: The "Loss-Plus-Threat" Recipe

After crunching the numbers from all these studies, the researchers found something fascinating. They discovered that there is no single "Jealousy Face" that pops up every time. Instead, jealousy seems to be a multichannel mix of two main things: Loss and Threat.

The "Loss" Ingredient (Sadness)
The most consistent clue they found was sadness. Whether looking at a frozen photo or a moving video, sadness showed up very clearly.

In the "Movie" (Dynamic): People feeling sad (or the loss part of jealousy) tend to look down, pull their bodies in, and slump their shoulders. It's a universal "I'm hurt" signal.
In the "Photo" (Static): Sadness also has a specific look in photos, often associated with a bluish tint to the skin and a softer, rounder face shape.
The researchers suggest that the "loss" part of jealousy is the most reliable part of the mix. If you see someone looking down and shrinking away, that's a strong signal of the hurt feeling at the core of jealousy.

The "Threat" Ingredient (Anger)
The second part of the mix is anger, which acts as the "threat" signal. This part is a bit trickier because anger often looks like other high-energy emotions (like disgust or fear).

In the "Photo" (Static): Anger is very strong in frozen images. The researchers found that when a face looks redder or has a sharper, more angular shape (like a square jaw), people immediately think "Threat!" or "Anger!" These are powerful, fast clues.
In the "Movie" (Dynamic): While anger has some movement clues, the study found that the "static" clues (the redness and sharp angles) were actually stronger and easier to spot than the moving clues for anger.

The "Recipe" for Jealousy
So, what does jealousy look like? The paper suggests it's not a new, weird emotion with its own face. It's a layered cake.

The bottom layer is the "Loss" signal (sadness): looking down, slumping, feeling hurt.
The top layer is the "Threat" signal (anger): a redder face, a sharper jaw, a look of defensiveness.

The researchers call this the "Loss-Plus-Threat" architecture. It means that when someone is jealous, they are likely showing a mix of "I am losing this person" (sadness) and "I am ready to fight for them" (anger). The sadness part is the most consistent and clear signal across both photos and videos, while the anger part is mostly visible in the quick, static snapshots of a face.

What This Means (And What It Doesn't)

The paper is very careful to say what they didn't find. They explicitly ruled out the idea that there is one special, unique "jealousy face" that is totally different from anger or sadness. You won't find a specific "jealousy nose" or "jealousy eyebrow" that appears only when someone is jealous. Instead, jealousy is a combination of other feelings we already know how to read.

The researchers are also honest about the limits of their detective work. They noted that none of the nine studies they looked at actually asked people to act out "jealousy" and then measured their faces. Instead, they looked at studies of anger, sadness, and distress and said, "Okay, if jealousy is made of these, then this is what it probably looks like." This is a smart way to guess, but it's still a guess. The paper suggests that future scientists need to go out and actually film people feeling jealous to see if this "Loss-Plus-Threat" recipe holds up in real life.

Also, the study pointed out that most of the people in these experiments were from Western, educated, industrialized, rich, and democratic (WEIRD) backgrounds. We don't know yet if a "jealous face" looks the same in every culture, or if different cultures have different rules for showing these feelings.

The Takeaway

In simple terms, this paper tells us that jealousy is a complex, two-part emotion. It's not a single, weird expression. It's a mix of sadness (the feeling of losing someone) and anger (the feeling of being threatened).

If you see someone looking down and slumped, they are likely feeling the loss part of jealousy.
If you see someone with a red face and a sharp, angry look, they are likely feeling the threat part.
When you see both at the same time, you might be looking at jealousy.

The study gives us a new way to think about jealousy: not as a monster with a scary face, but as a human reaction that combines the pain of loss with the fire of defense. It's a "loss-plus-threat" signal that our brains are constantly trying to decode, using both the frozen snapshots of our faces and the moving stories of our bodies.

Technical Summary: Jealousy Without a Face

Problem Statement
Jealousy is theoretically conceptualized as a complex, multicomponent emotional episode arising from the threat of losing a valued relationship to a rival. While its cognitive appraisals and self-reported experiences are well-documented, its involuntary nonverbal manifestations remain empirically underdeveloped. A central gap exists in the literature: there are nearly no primary studies that treat jealousy itself as a nonverbal dependent variable in adult populations. Consequently, it remains unclear whether jealousy possesses a unique nonverbal "signature" or if it is expressed through a configuration of component emotions, specifically anger, sadness, and distress. This study addresses this gap by systematically reviewing and meta-analyzing experimental evidence on the nonverbal expressions of these theoretically relevant component emotions to infer the expressive architecture of jealousy.

Methodology
The authors conducted a systematic review and meta-analysis following PRISMA guidelines, with the protocol preregistered on the Open Science Framework (OSF).

Search Strategy: Searches were conducted in PubMed, PsycINFO, Scopus, and Web of Science (2000–2025). Due to the scarcity of studies directly linking jealousy to nonverbal expression, the search strategy was expanded to include the component emotions of anger, sadness, and distress, grounded in multicomponent theories of jealousy.
Eligibility: The review focused on human adult samples (non-clinical) in experimental or observational settings where nonverbal expressions (facial, behavioral, physiological) were the dependent variable. Studies relying solely on self-reports, clinical populations, or lacking full text were excluded.
Selection: From 1,073 identified records, 9 studies (yielding 10 independent effects) were included in the final synthesis, with a total analyzed sample of approximately 1,847 adults. One study was excluded due to a serious risk of bias in confounding.
Risk of Bias: Study quality was assessed using the ROBINS-I tool across seven domains. Inter-rater reliability was moderate ( $\kappa = .59$ ). Only one study achieved a "Low" overall risk of bias; the remainder were rated "Moderate."
Analysis: A random-effects model (DerSimonian & Laird) was applied to synthesize effect sizes. Key statistics (e.g., $\eta^2$ , $F$ , $t$ , $OR$, $R^2$ ) from the included studies were converted to Cohen's $d$ . Heterogeneity was assessed using Cochran's $Q$ and $I^2$ .

Key Results

Pooled Effect Size: The meta-analysis yielded a large, robust pooled effect size ( $d = 1.17$ , 95% CI [0.87, 1.47]), indicating that nonverbal signals associated with jealousy-related emotions are detectable and multimodal.
Signal Types: The analysis distinguished between static signals (structural/chromatic cues from single moments, e.g., facial color, shape, posed configurations) and dynamic signals (temporal/movement cues, e.g., gaze shifts, head tilts, micromovements).
- Static Signals: These produced the largest individual effect sizes. Skin color manipulations (specifically increased redness associated with anger and greenish-blue shifts with sadness) showed the strongest effects ( $d \approx 1.72–1.99$ ). Facial angularity and specific Action Units (e.g., brow lowering/AU4 for anger) also demonstrated strong predictive power for threat-related states.
- Dynamic Signals: These produced statistically significant but generally smaller effect sizes ( $d \approx 0.61–0.79$ ). Their primary contribution was indexing expressive authenticity, motivational state, and the temporal unfolding of emotion.
Component Specificity:
- Sadness: Emerged as the most distinctive component, exhibiting a coherent multichannel signature across both static (color, configuration) and dynamic (downward gaze, withdrawal posture) domains.
- Anger/Threat: Primarily anchored in high-magnitude static cues (redness, angularity) but showed less distinctiveness in dynamic patterning due to overlap with other high-arousal negative states.

Key Contributions

Identification of an Expressive Architecture: The paper identifies a "candidate expressive architecture" rather than a validated, unique nonverbal signature of jealousy. It proposes a "loss-plus-threat" model where jealousy is expressed through a composite of sadness-like loss signals (forming a multichannel backbone) overlaid with static threat markers.
Clarification of the Empirical Gap: The review explicitly highlights the absence of studies treating jealousy as a nonverbal dependent variable in adults, contrasting this with the robust evidence for its component emotions.
Differentiation of Signal Channels: The study provides empirical evidence that static and dynamic signals are complementary rather than interchangeable. Static cues offer rapid, high-magnitude categorization of threat vs. loss, while dynamic cues refine these inferences by signaling authenticity and motivational intent.

Significance and Claims
The authors claim that their synthesis provides an empirically grounded foundation for future research, moving the field from inference to direct observation. They argue that jealousy should be conceptualized not as a discrete basic emotion with a unique facial prototype, but as a multicomponent configuration organized around relational loss (sadness) with context-sensitive layers of threat (anger).

The paper modestly concludes that while the "loss-plus-threat" architecture offers a heuristic template for future observational paradigms, the current evidence is inferred from component emotions rather than observed directly in jealousy episodes. The authors emphasize that their findings do not prove the existence of a specific "jealousy face" but rather map the building blocks (anger, sadness, distress) from which jealousy inferences may be constructed. They call for future preregistered, multimodal studies in ecologically valid contexts to directly test these proposed configurations, noting that current limitations include the lack of direct jealousy operationalization, the restriction to WEIRD populations, and high heterogeneity in measurement methods.

The Detective Work: How They Solved the Mystery

The Big Findings: The "Loss-Plus-Threat" Recipe

What This Means (And What It Doesn't)

The Takeaway

Technical Summary: Jealousy Without a Face

More like this