One day, while scrolling through Instagram, I noticed that body display seemed to make up a larger share of recommended content. Some posts were fitness records, some were fashion, and some were difficult to place in only one category.

▶ Earlier audio summary (not yet updated for this revision)
AI-synthesized voice, cloned from the author's own voice

That is my own viewing experience; it cannot stand in for the whole platform. What I want to ask is this: what relationship do those recommendations have with my interactions? When creators see that one kind of image receives more responses, how might they adjust what they post next?

Treating body display as a kind of “language” is how I make sense of this interaction. Before discussing the metaphor, though, we need to distinguish what recommendation systems actually say they do.

Drawing attention does not necessarily mean receiving recommendations

In its explanation of recommendation ranking, Meta says its systems combine many predictions, including whether a user might share content and survey feedback signals. Recommendations therefore cannot be reduced to a single contest for viewing time.

Another condition that is easily missed is recommendation eligibility. Meta’s sensitive-content explanation distinguishes between content being allowed to exist and being suitable for recommendation: some sexually suggestive content may not violate community rules but can still face recommendation limits.

So “the more nude, the higher the reach” is not a general rule supported by these documents. The performance of an account has to be considered with its audience, form of content, interactions, and platform rules. My own feed cannot by itself answer which factor produced a particular recommendation.

What do I mean by “body language”?

Fitness, fashion, dance, and artistic creation can all express different things through the body. We cannot decide that a creator only wants traffic simply because more skin is visible.

I call composition, pose, clothing, and the relationship to text the “grammar” of this language. It is an analytical metaphor: the same photograph can be understood differently when placed with different text or in a different community.

The metaphor cannot yield a fixed scale of nudity, nor is it enough to divide East Asian and Western audiences into two preferences. Claims about which forms work better in which place need clear samples and comparison.

Responding to numbers may gradually change creation

Imagine a creator finds that one type of body photograph receives more interaction and posts more of it; viewers continue to respond, and the creator keeps adjusting. This is a possible cycle I want to discuss, not a case I have tracked and verified.

The adjustment is not necessarily a problem. Creators naturally respond to audiences, and audience preference can be part of creation. What deserves attention is whether reach gradually becomes the only criterion, making it difficult to distinguish what the creator wants to express from the boundaries they do not wish to surrender.

In “Technology Begins with Humanity: Business Lessons from Facebook’s Algorithmic Reconstruction”, I discussed how platforms influence business choices. With body display, there is another personal question: is the manner of being seen still one the person is willing to accept?

Viewers also participate in the process

Viewers’ watching and interaction can become signals a platform can read, but a signal does not fully explain a person’s feelings. Someone may linger out of appreciation or curiosity, or because they are uncomfortable; numbers alone cannot tell all those meanings.

Likewise, when creators see the results of a post, they do not thereby know why every person watched. Treating interaction counts directly as others’ evaluation of one’s body puts too much meaning into a single number.

I want to retain this distinction: platforms record behavior; human worth and the meaning of creation still need human interpretation.

Keep a way of speaking that one is willing to use

For creators, it may help to ask before publishing: What does this post want to express? How much am I willing to make public? If the response is poor, do I still think it is worth keeping? These questions add another basis to evaluating results, so the decision does not rest only on one reach number.

Viewers can also ask: Why am I continuing to watch? Is this content I chose actively, or merely the next post that happened to appear? We do not need to turn every scroll into self-judgment, but we can practice recognizing our choices.

Body display can be creation, work, documentation, or simply liking a photograph. What I want to protect is room for the people involved to explain those meanings, rather than having them all replaced by the phrase “the algorithm likes it.”