The graphic is orange. That is the first fact, and it is not a small one. On a garment whose retail title includes the word “costume,” the pumpkin face arrives as a flat, saturated orange shape against the shirt ground, not as a printed photograph of a gourd, not as a gradient, not as a fabric texture. The color is doing the work of a category: this is a pumpkin because it is orange and because the silhouette around it is round. The rest of the design is a set of decisions about what kind of pumpkin it will be.

The eyes are round. Not triangular, not slanted, not narrowed. Round eyes on a carved face read as surprise or as innocence, depending on the mouth. The nose is heart-shaped. That is the detail that reorganizes the whole image. A heart-shaped nose is not a standard carving pattern; it is a choice that pulls the face away from the scary end of the jack-o’-lantern tradition and toward the cute end. The mouth is smiling. Together, the round eyes, the heart nose, and the smile produce a face that is legible at a glance as friendly. The retail title says “Cute Kids,” and the artwork earns that adjective through the nose more than through anything else.

This is the variant that matters. A jack-o’-lantern face is a template with a narrow set of variables: eye shape, nose shape, mouth shape, and the overall proportion of the carved area to the pumpkin body. Change the nose from a triangle to a heart and you have changed the emotional register of the object. The shirt is not a generic pumpkin; it is a specific carving decision, and the decision is the design.

The phrase “jack o lantern shirt” describes a category of garment that appears in search data with a reported monthly range of 100 to 1,000 in the United States, per Google Keyword Planner figures extracted on 2026-10-12 for the period September 2025 through August 2026. That range is a demand signal, not a sales figure, and it does not tell us anything about this particular shirt. What it tells us is that the category is active and that the shirt is competing inside a recognizable type. The design has to distinguish itself within the type, and the heart nose is how it does that.

The shirt is listed as a standard t-shirt, and the artwork is described in the retail observation as an orange jack-o’-lantern face graphic with round eyes, a heart-shaped nose, and a smiling mouth. That is the full visual record available. There is no claim here about fabric weight, fit, or print method, because those are not in the observation. The design is the subject.

What makes the cute variant work as a printed object is its economy. The face is built from a small number of shapes: two circles, one heart, one curved smile, and the implied round boundary of the pumpkin. There is no stem, no vine, no background, no text. The retail title includes the word “costume,” but the graphic is not a costume in the sense of a full-body garment; it is a face printed on a shirt, and the wearer supplies the body. The shirt is a costume in the way a mask is a costume: it gives the viewer a face to read.

That reading is where the design’s argument sits. A scary jack-o’-lantern face is a warning; a cute one is an invitation. The heart nose is the invitation. It says the pumpkin is happy to see you. The round eyes say the pumpkin is not hiding anything. The smile says the encounter is benign. On a shirt, that combination functions as a social signal: the wearer is participating in Halloween without committing to menace. The design is not neutral; it is choosing a side in the long argument about what Halloween is for.

The shirt is available at the Amazon listing, where the title identifies it as a kids’ Halloween costume pumpkin design from Quirky Cotton Collective. The listing is the source for the object’s identity; the visual observation is the source for the artwork’s description. Neither source establishes a wearing experience, a customer reaction, or a publication history, and none is claimed here.

Editorial note: This publication and Quirky Cotton Collective share an owner. This is not an independent review.

The useful question about a shirt like this is not whether it is cute. The artwork answers that. The useful question is what the cuteness is doing. A heart-shaped nose on a carved pumpkin is a small act of design discipline: it takes a form associated with threat and makes it safe without making it bland. The round eyes and the smile do the same work. The result is a face that can be worn by a child without requiring the child to perform fright. That is a specific achievement, and it is visible in the orange shape on the shirt.

The graphic is a carved pumpkin face, rendered in orange, and it is doing two things at once. The left eye is a closed crescent: a wink. The mouth is a wide, uneven grin with a tongue hanging out of it, drawn as a separate rounded shape below the lip line. The whole face sits inside a roughly circular pumpkin silhouette. There is no stem visible in the supplied observation, no vine, no background field, no lettering. The design is the face and nothing else.

That absence is the first thing worth noting. A jack-o’-lantern is a carved object, and carving is subtractive: you remove material to make the face. The shirt reproduces the result of that removal as a flat printed image. The wink and the tongue are not carved features in any literal sense — a tongue is not something you cut out of a pumpkin — but they read as if they were, because the whole composition borrows the visual grammar of a carved gourd: triangular or crescent eye shapes, a jagged or scalloped mouth, a face that occupies the full available surface. The design is a costume in print form, and the goofy variant is a specific choice within that grammar.

Consider what the wink does to the face’s address. A carved pumpkin face is usually a stare: two eyes, a mouth, aimed outward. The wink collapses one eye into a line, which changes the symmetry and, more importantly, changes the timing. A stare is static. A wink is an event — it happens, and then it is over, except that a printed wink never stops happening. The shirt freezes a gesture that in a living face lasts a fraction of a second. That is the comic engine of the design: it is a permanent wink, a joke that cannot land and then move on. The tongue reinforces it. A tongue sticking out is a second gesture layered on the first, and the two together produce a face that is not scary, not cute, but deliberately foolish. The retail title calls it “goofy,” and the visual observation supports that word: the wink and the tongue are the two features that make the face read as a joke rather than a threat.

This is where the shirt type matters. The jack-o’-lantern carved face costume is a genre with its own conventions: the wearer becomes the pumpkin, the face is large and centered, and the design does the work of a mask without the mask. The goofy variant is a deviation from the genre’s default severity. A scary carved face — the jagged mouth, the narrowed eyes — signals menace. The goofy variant signals that the wearer is in on the joke. The wink is the tell. It is the difference between a pumpkin that is watching you and a pumpkin that is kidding around.

The design also has a compositional problem to solve, and it solves it by omission. A pumpkin face printed on a shirt has no edges to work with: no rind, no ribs, no shadow. The supplied observation describes an orange face graphic, not a pumpkin body. So the face must carry the whole object. It does this by keeping the mouth wide and the eyes high, so the silhouette reads as round even without a drawn circle. The tongue, hanging below the mouth line, extends the face downward and gives the composition a bottom edge it would otherwise lack. Remove the tongue and the face would float; with it, the face has a weight and a floor. That is a small piece of design logic, but it is the kind of thing that separates a printed costume from a printed sticker.

The shirt is sold under the title “Goofy Jack O Lantern Face Halloween Costume Pumpkin” by Quirky Cotton Collective. The phrase “jack o lantern shirt” describes the object a reader might be looking for, and this is one: a shirt that turns the wearer into a carved face for a night. The design does not depend on lettering, so there is no caption to read and no gap between picture and words. The joke is entirely visual, which makes it portable across a group: several people in the same shirt become a row of identical winks, and the repetition is the point. A group costume works when the individual unit is legible at a distance and funny up close. The wink is legible at a distance; the tongue is the up-close reward.

What the design does not do is claim a history. It does not cite a particular carving tradition, a region, or a folk practice. It is a graphic in the shape of a pumpkin face, and its references are to the costume aisle rather than to any archive. That is not a criticism; it is a description of the object’s ambition. The shirt is a disposable visual joke, and the wink is its whole argument: the face is not trying to frighten anyone, and it is not trying to be beautiful. It is trying to be the pumpkin that makes the group photo work.

Editorial note: this publication and Quirky Cotton Collective share an owner.

The shirt is available at the Amazon listing.

The Verbal Contract: What Tino Sehgal’s Oral Sale Actually Records

The paperwork problem with Tino Sehgal is not that there is none. It is that the paperwork has been replaced by a room. In a 2006 interview with The Art Newspaper, Sehgal described the sale of his work in plain terms: “Instead of a paper certificate, there’s an oral sale contract.” That sentence is usually quoted as a provocation. Read forensically, it is a description of an instrument — one with parties, witnesses, conditions, and a payment clause. The interesting question is not whether the contract exists. It is what the contract records, who is present when it is made, and what happens when the people in the room leave.

The instrument and its witnesses

According to the same interview, the oral contract contains “three to six conditions depending on the work,” and a lawyer or notary is present. A curator or art critic normally serves as witness, alongside the buyer and Sehgal’s gallerist. That is a quorum, not an absence. The document has been dissolved into a procedure: conditions recited, a legal officer present, a witness drawn from the professional field, the counterparties in the room.

This matters because the standard complaint about Sehgal — that nothing is written down — misdescribes the governance. A written contract is a record that survives the meeting. An oral contract is a record that survives only as long as the people who heard it, and the institution that employs them, continue to treat it as binding. The sale does not eliminate the record. It relocates the record from an archive to a room of named people, and then bets on their tenure.

The money clause

Sehgal told the interviewer he accepts “cash and wire, but we don’t do cheques,” with a stated reason: “I don’t want to have one material thing that solely represents a transaction.” This is the clause most often repeated as evidence of purity. It is more useful read as a drafting choice. A cheque is a single negotiable instrument that stands for the deal; a wire transfer is a bank record that stands for the deal plus a chain of intermediaries. The rule does not eliminate paper. It eliminates one specific artifact — the cheque — while accepting another. The distinction tells us what counts as residue in this system: not records as such, but records that could be mistaken for the work’s certificate.

That is a narrow and coherent position. It is also the kind of position that only makes sense if the sale is understood as transferring something other than a document. Which raises the next clause.

The edition clause

Sehgal stated the works are in an edition of four, and that private individuals can buy them as well as institutions. An edition of four is a familiar market structure applied to an unfamiliar object. It limits supply, it creates a set of holders who can be identified, and it gives the artist a defined number of transmission relationships to maintain. The edition is the closest thing to a registry in the entire arrangement — not a written registry, but a countable one.

The private-buyer provision is the part that complicates the institutional story. If a work can be sold to an individual, then the transmission chain depends on a household rather than a department. The interview does not describe how a private owner is expected to maintain the conditions, and no retrieved source supplies that answer. What the record does establish is that the sale is not restricted to museums, whatever the emphasis on institutional buyers suggests.

The continuity clause, which is really a hiring criterion

The most consequential passage in the 2006 interview concerns staff turnover. Asked whether the work must be relearned when the people who witnessed the contract leave the institution, Sehgal answered: “Exactly — so we try to find people we think are going to stay for a while in an institution!”

Read that as a contract term and it is unusual. The buyer’s identity is partly selected for expected tenure. The sale is underwriting a transmission chain, not merely transferring title. A museum that acquires a Sehgal is being assessed not only on its ability to pay but on its ability to retain the staff who will carry the conditions forward. That is a governance choice presented in the vocabulary of conservation. It also means the oral contract has an implicit expiry risk that a written contract does not: the record degrades with every resignation.

The 2006 interview also records that Tate had acquired This is propaganda the previous year. The acquisition is stated; the acquisition file is not. That gap is the honest limit of this essay. What the interview establishes is that a major institution bought a work whose terms were, by the artist’s account, conveyed orally. What it does not establish is what Tate’s registrar wrote down, if anything, and no retrieved source answers that.

The artist’s own framing, twenty years on

In a 2026 interview with The Korea Times, Sehgal described the work in terms that are more useful to a forensic reading than the familiar “art without objects” line. “My works are commodities,” he said. “But they are immaterial commodities.” He also said: “My work can’t really be destroyed, but it can be forgotten.” And he described a studio team that transmits the pieces to museums and new generations of interpreters, compared to choreographic trusts that preserve a choreographer’s repertory.

Those three statements, taken together, describe the actual risk profile. The work is not vulnerable to physical loss. It is vulnerable to institutional forgetting — to the departure of the people who heard the conditions, the loss of interest of the institution that holds the edition, the failure of the studio team to transmit. The oral contract is a transmission mechanism, and like any transmission mechanism it has failure modes that are organizational rather than material.

What the record does not contain

No retrieved source provides the text of any Sehgal condition, the identity of any witness, the terms of any resale, or the content of any museum acquisition file. No court filing describing a dispute over a Sehgal oral contract was retrieved. No legal analysis of enforceability in a particular jurisdiction was retrieved. Those absences should be stated rather than filled. The claim that the oral contract is legally enforceable in any specific sense is not supported by the material available here, and the claim that it is unprecedented in art history is likewise unsupported.

What the record does support is narrower and more interesting. A sale took place with a lawyer or notary present, a professional witness, a set of recited conditions, a payment method chosen to avoid a specific artifact, an edition of four, and a buyer selected partly for staff longevity. That is a contract. It is simply a contract whose archive is a room, and whose enforcement depends on the people in the room staying long enough to remember what they heard.

FAQ

Does Tino Sehgal really sell work without any written contract?

By his own account in a 2006 interview with The Art Newspaper, the sale proceeds by an oral contract rather than a paper certificate, with three to six conditions depending on the work and a lawyer or notary present. The interview does not establish that no document of any kind is created; it establishes that the certificate is oral.

Who witnesses a Sehgal sale?

According to the same interview, a curator or art critic normally serves as witness, alongside the buyer and Sehgal’s gallerist, with a lawyer or notary present.

How is a Sehgal work paid for?

Sehgal stated that he accepts cash and wire transfers but not cheques, with the stated reason that he does not want a single material object solely representing the transaction.

How many editions of a Sehgal work exist?

He stated the works are in an edition of four, and that private individuals as well as institutions can buy them.

What happens when museum staff who witnessed the contract leave?

Sehgal acknowledged that the work must be relearned by new people, and said buyers are chosen partly for how long they are expected to stay at an institution. The 2026 Korea Times interview describes a studio team that transmits the pieces to museums and new generations of interpreters.

Can a Sehgal work be destroyed?

Sehgal stated in 2026 that the work cannot really be destroyed but can be forgotten. The risk he names is institutional memory, not physical loss.

The shirt reads, in two lines of distressed capital letters, TWEEDLE DUM above I’M WITH DEE, and beneath the second line a solid arrow points to the right. That is the whole design. There is no illustration of a round little man in a propeller beanie, no checkerboard, no mushroom, no rabbit. The object is a caption and a directional mark, and the caption only works if another person is standing to the wearer’s right.

This is the Dum half of a pair. The companion shirt, which I have written about separately, carries the Dee name and an arrow pointing left. Set the two wearers side by side in the correct order and the arrows point at each other; the phrases resolve into a single sentence distributed across two torsos. Set them in the wrong order and the arrows point outward, away from the couple, and the joke inverts into two people each insisting on their own identity while gesturing at empty room. The design has a correct installation, like a diptych, and the wearers are the ones who have to hang it properly.

The lettering is the first thing to notice after the words themselves. It is not clean type. The edges of the capitals are broken and flecked, the fill uneven, as though the text had been stamped through a worn screen or photocopied until the toner gave out. This is a deliberate graphic choice and it does specific work: it makes the phrase look like a label rather than a slogan, something applied to a surface rather than spoken by a person. A crisp sans-serif would read as merchandise. The distressed face reads as a marking, and markings are what you put on things you are trying to keep track of.

That is the argument I want to make about this shirt. It is not a costume. It is an inventory tag that a person has agreed to wear, and the arrow is the part that makes the agreement visible.

Consider what the arrow does that the words cannot. I’m with Dee is a claim about a relationship, but it is a claim made to no one in particular; it could be true of an absent friend, a person across the room, a memory. The arrow converts the claim into a vector. It specifies a direction and therefore a location, and a location implies a body that should be occupying it. The shirt does not say I am with Dee in the way a name badge says Hello, my name is. It says it the way a shipping label says this side up: as an instruction to whoever is handling the object, telling them where the rest of the shipment has gone.

This is why the paired costume is a more interesting form than the solo one. A single novelty shirt asks the wearer to perform a character. A matched pair asks two wearers to perform a system — a relationship with a correct configuration and a failure mode. The failure mode is the whole comic engine. Tweedle Dee and Tweedle Dum are, in the source text, indistinguishable except by name, and the names are the only thing that tells you which is which. The shirts reproduce that problem exactly: the two garments are identical in layout, identical in lettering, identical in distress, and differ only in the words and the direction of the arrow. You cannot tell the halves apart without reading them, and reading them requires standing close enough to a stranger’s chest to be socially awkward. The design has built the confusion of the characters into the logistics of the costume.

The arrow, in other words, is doing the work that the names cannot. It is a redundancy measure. If the words fail — if the light is bad, if the wearer is turned, if the viewer does not know the poem — the arrow still points. It is the same principle as the colored dots on a wiring harness: the label tells you what the wire is, and the color tells you again, because the label will eventually be unreadable and the wire will still need to go somewhere.

I am aware that this is a heavy reading of a Halloween shirt. But the heavy reading is the correct one, because the shirt is not asking to be taken lightly. It has chosen a font that looks like a stencil, a phrase that functions as a caption, and a directional mark that only makes sense in the presence of a second object. Every one of those choices is a choice about legibility under conditions of separation. The design assumes that the two halves will spend most of their lives apart — in a drawer, in a closet, on two different people who arrived at the party separately — and it is built to survive that separation and re-pair correctly when the halves are reunited. That is not a costume. That is a labeling protocol.

The distressed lettering reinforces the point. Clean type would suggest a printed message, something issued. Worn type suggests something that has been through handling — that has been packed, shipped, unpacked, and read under poor light. The shirt is presenting itself as an object with a history of transit, even though it is new. This is a small fiction, but it is the same fiction that makes a shipping crate stencil legible: the wear on the letters is evidence that the marking has been used, and a marking that has been used is a marking you can trust. The design borrows the authority of the loading dock.

What the shirt does not do is resolve the question of which half is which. The source material gives the characters no distinguishing feature beyond their names, and the shirt follows suit. There is no visual cue — no color code, no icon, no difference in the arrow beyond its direction — that would let a viewer identify the Dum half without reading the words. This is either a flaw or the entire point, and I think it is the point. The paired costume is funny because the pairing is arbitrary. Two people put on two nearly identical shirts and agree, for one evening, to be a matched set. The agreement is the costume. The shirts are just the paperwork.

The shirt is sold as the Tweedle Dum I’m With Dee matching Halloween costume, and the listing treats it as one half of a set. That framing is accurate to the object. A single shirt of this design, worn alone, is a fragment: an arrow pointing at nothing, a caption with no referent. The design only closes when the second half arrives.

Editorial note: this publication and Quirky Cotton Collective share an owner. This is not an independent review.

The system the shirt makes possible is small and specific: two people, one evening, a correct arrangement of arrows. It is not a profound system, but it is a legible one, and legibility is rarer than it looks. Most matching costumes ask the wearer to look like something. This one asks the wearer to be findable — to stand in the correct position, point in the correct direction, and let the label do the rest. The shirt is a way of saying I am here, and the other one is over there, in a font that looks like it has been through a warehouse. That is a modest achievement. It is also, on the evidence of the design, exactly what was intended.

The first thing the shirt does is refuse a handshake. A yellow caution triangle sits at the top of the print, the kind of triangle that on a loading dock means wet floor or high voltage or authorized personnel only. Below it, a hand is crossed out with a diagonal bar. Below that, in warning-sign lettering, the words: NOT FRIENDLY DO NOT TOUCH.

The design is a public notice worn on a body. That is the whole of it, and it is enough to make the object worth reading closely. The shirt does not say introvert. It does not say anxious or tired or please text first. It borrows the grammar of industrial safety — triangle, pictogram, prohibition bar, all-caps imperative — and applies that grammar to a person. The joke is not that the wearer is unfriendly. The joke is that unfriendliness has been formatted as a hazard.

Look at the order of operations. The triangle comes first, which is the position of maximum visual authority on any sign. A caution triangle is not a suggestion; it is a classification. It tells the reader that what follows is a recognized category of risk, one that has been assessed and labeled by someone with the authority to label it. The crossed-out hand comes second, which is the position of the specific prohibition. The words come third, which is the position of the redundant explanation — the part of the sign that exists because people ignore pictograms. The design is built like a real sign, and that is why it works. It does not ask the wearer to perform hostility. It asks the wearer to stand behind a notice.

The crossed-out hand is the most interesting element, because it is doing two jobs at once. In standard signage, a crossed-out hand means do not touch the equipment. On a shirt, it means do not touch the person. The pictogram does not change; the surface it is printed on changes what the pictogram refers to. This is the small mechanical trick at the center of the design, and it is worth naming precisely: the shirt takes a symbol whose meaning is fixed by industrial convention and relocates it onto a body, where the same symbol now reads as a boundary. The caution triangle does not warn about the shirt. It warns about the wearer. The wearer has become the equipment.

There is a comic timing in the lettering that the retail description does not capture. NOT FRIENDLY is a statement of condition. DO NOT TOUCH is an instruction. The two are set in the same warning-sign type, at the same scale, with no punctuation between them. The absence of a period or a comma matters. A period would make the two phrases sequential — first a diagnosis, then a rule. Without it, the phrases run together as a single compound warning, the way a real sign reads when it has been designed by a committee that wanted to cover every liability. The shirt is funnier for being badly punctuated in the way that actual safety signage is badly punctuated. It sounds like it was drafted by someone in a fluorescent vest who had one line of budget for words.

The phrase not friendly is also doing quieter work than it first appears. It is not hostile. It is not dangerous. It is the language of a customer-service note, the kind of thing a receptionist might say about a visitor who has been standing too long by the desk. Not friendly is a classification, not an insult. It places the wearer in a category that is legible to institutions — the person who does not want to be greeted, the person who has already decided not to participate in the small talk. The shirt is not a threat. It is a status. The caution triangle elevates that status to a hazard, and the joke lives in the gap between the mildness of not friendly and the severity of the triangle.

I have written before about shirts that use a single punctuation mark to carry the whole joke. This one uses a different mechanism: it uses a borrowed format. The warning sign is a genre with its own rules, and the shirt follows those rules closely enough that the reader recognizes the genre before reading the words. That recognition is the setup. The punchline is the application of the genre to a person. The design does not need a caption or a slogan or a second line of text. The format is the joke.

What the shirt does not do is resolve the tension it creates. A real warning sign is meant to be obeyed. A shirt that says DO NOT TOUCH is meant to be read, and then the reader has to decide what to do with it. The design leaves that decision open. It does not say ask first. It does not say I am joking. It says what it says, in the format that says it most clearly, and then it stops. That is the discipline of the object. It trusts the reader to understand that a warning sign on a body is a performance of a boundary, not a legal notice. The shirt is not actually a hazard. It is a shirt about the desire to be treated as one.

The Not Friendly Do Not Touch shirt is listed by Quirky Cotton Collective. The design is a warning-sign composition: a yellow caution triangle, a crossed-out hand, and the words NOT FRIENDLY DO NOT TOUCH in the lettering of a public notice. The shirt is a standard t-shirt. The print is the argument.

There is a practical question that the design raises and does not answer: what does the wearer do when someone ignores the sign? The shirt has no escalation. It has no second panel, no follow-up instruction, no phone number to call. It is a single-format object, and its authority depends on the reader accepting the format. In a workplace, that acceptance is not guaranteed. The caution triangle is a familiar shape, but the person wearing it is not a piece of equipment, and the reader knows it. The shirt works only if the reader is willing to play along with the fiction that the wearer has been officially classified as a hazard. That willingness is the social contract the design depends on, and it is not printed anywhere on the shirt.

The design also has a group-costume life that is worth noting without overclaiming. A warning sign is a modular format. It can be reproduced at different scales, in different colors, with different pictograms. A group of people wearing the same warning-sign shirt becomes a set of identical notices, which is a different visual effect from a single person wearing one. The shirt does not require that reading, but the format makes it available. The caution triangle is a unit, and units can be repeated.

The object is not a document. It is a shirt. But it borrows the authority of a document — the triangle, the bar, the all-caps imperative — and it asks the reader to treat that authority as real for as long as it takes to read the words. The joke is not that the wearer is unfriendly. The joke is that unfriendliness has been given a format, and the format is legible. The shirt is a small piece of signage that has been detached from its wall and attached to a person, and the person is now the wall. That is the whole of the design, and it is enough.

Editorial note: This publication and Quirky Cotton Collective share an owner.

The graphic is an orange pumpkin face. Two triangular eyes sit above a jagged mouth, the kind of zigzag cut a carver makes when the knife changes direction at every tooth. The orange is the only color in the design; the face is the whole composition. There is no stem, no vine, no field, no moon. The shirt is a jack-o-lantern and nothing else.

That restraint is the argument. A costume t-shirt has to do in one printed plane what a carved pumpkin does in three dimensions: it has to read as a face at conversational distance, and as a scary face at the distance of a doorway. The triangular eyes are the load-bearing element. They are wide at the base and narrow at the top, which is the geometry of an angry brow, not a surprised one. The jagged mouth is the second element. Its teeth are irregular, which is the visual shorthand for damage — a mouth that has been broken rather than drawn. The two together produce the expression the retail title calls “scary.” I am reading the supplied visual observation, not a historical claim about pumpkin carving.

What interests me is the gap between the object and its category. The listing calls this a jack o lantern shirt, and the phrase is accurate in the way a shipping label is accurate: it tells you what bin the object belongs in, not what the object does. The shirt does not depict a pumpkin. It is the pumpkin, worn on the torso, with the face where the face goes. That is a costume logic, not a graphic-tee logic. A graphic tee puts an image on a body; a costume tee substitutes the body for the image. The wearer’s head becomes the stem. The wearer’s arms become the vine. The shirt is the carved surface.

This is why the design has no text. Several sellers in this genre use a no-text graphic, and the choice is structural rather than stylistic. Text would break the substitution. A caption — “Scary Pumpkin,” “Happy Halloween,” a pun — would pull the shirt back into the category of printed commentary, where the wearer is a person wearing a joke. Without text, the wearer is a pumpkin with a person inside it. The jagged mouth is the only sentence the shirt speaks, and it speaks it in teeth.

The composition also solves a problem that costume tees usually fail: the neckline. A printed face on a t-shirt has to negotiate the collar, which interrupts any design that runs too high. Here the triangular eyes sit below the collar line, and the jagged mouth sits at the center of the chest. The face is complete within the printable area. Nothing is cropped, nothing is implied. The shirt does not ask the wearer to complete the image with their own face; it asks the wearer to replace their face with the printed one. That is a more demanding ask, and the design earns it by keeping the face whole.

The jagged mouth is also the element that dates the design least. A carved pumpkin’s mouth changes with every carver; the zigzag is a technique, not a trademark. The shirt borrows the technique and stops there. It does not borrow a franchise, a film, or a licensed character. The orange is the orange of a pumpkin, not the orange of a brand. That makes the shirt reusable across Octobers in a way that licensed costume tees are not, and it makes the design legible to anyone who has ever seen a carved pumpkin without requiring them to have seen a particular movie.

The retail title uses the word “costume,” and the word is doing real work. A costume is a temporary identity, put on and taken off. The shirt is a costume that can be worn to a party and then worn again to a grocery store, where the jagged mouth reads as a graphic rather than a disguise. That double life is the shirt’s actual function. It is not a pumpkin costume that happens to be a t-shirt; it is a t-shirt that happens to be a pumpkin costume, and the difference is the jagged mouth. A smiling pumpkin is decoration. A jagged mouth is a threat, and a threat is what makes a costume read as a costume rather than as seasonal apparel.

The design’s weakness is the same as its strength: the face is the whole composition, and the face is small. At a distance the triangular eyes and jagged mouth resolve into a single orange shape, and the shirt becomes a pumpkin-colored rectangle. That is a real limitation for a costume, which is usually judged at a distance. The design compensates by keeping the face centered and the orange saturated, but it cannot compensate for scale. A wearer who wants the costume to read across a room is asking the graphic to do something the graphic was not built to do. The shirt is a close-range costume, and close-range costumes are the ones that get photographed.

I keep returning to the jagged mouth because it is the only element that carries risk. The triangular eyes are a convention; every carved pumpkin has them. The jagged mouth is a decision. A carver who cuts a jagged mouth is choosing asymmetry over neatness, damage over delight. The shirt makes the same choice, and the choice is what separates this design from the smiling-pumpkin shirts that share its category. The smile is the default. The jagged mouth is the edit.

Editorial note: This publication and Quirky Cotton Collective share an owner.

The Document That Tells the Story

On May 6, 2002, Larisa Muravska delivered a report to the Open Society Institute’s Arts and Culture Network Program. Its title was unassuming: Assessment/mapping of activities of the Soros Centers for Contemporary Arts. It ran to twenty-three pages plus attachments, and it was commissioned following a November 2001 decision of the program’s Sub-board. The report opens with a sentence that any forensic reader should underline: “Following the decision of its Sub-board, in November 2001, the Network Arts and Culture Program commissioned the assessment of the 19 former Soros Centers of Contemporary Arts (SCCAs).”

That number—19—is the first clue. The SCCA network had been described elsewhere as comprising twenty centers. The assessment covers nineteen. The discrepancy is not a scandal; it is a filing detail. But it is exactly the kind of detail that tells you what kind of document you are holding. This is not a celebration. It is an audit of a network that had already been spun off, and the foundation wanted to know what it had wrought.

The assessment is the primary evidence for how the Soros network built contemporary art infrastructure across Central and Eastern Europe. Not because it boasts of success—it does not—but because it records the machinery: budgets, staff counts, governance relations, sustainability targets, and the quiet admission of what it did not measure. Read as a governance document, it is more revealing than any press release.

The Institutional Timeline, From the Files

The C3 Center for Culture and Communication, successor to the Budapest SCCA, maintains a program history that reads like a chain of custody. The Soros Foundation Fine Arts Documentation Center was established in 1985 in the Budapest Mûcsarnok (Kunsthalle), as a cooperation between the Mûcsarnok and Soros Foundation Hungary. Its purpose, as the C3 page states, was to support modern Hungarian culture “which was banned or at least forced into the background by official cultural policies”—supporting artists in “counter-culture” circles through scholarships, purchases, exhibitions, and catalogues.

In 1991, the Center expanded under the name SCCA. From the end of 1991 until 1999, according to the Muravska assessment, “18 additional SCCAs were opened in 17 Central and Eastern European countries, with goals and activities similar to the Budapest SCCA: to promote, develop and support contemporary art.” The C3 page confirms the count: “From the end of 1991 until 1999, 19 additional SCCAs were opened in 17 Central and Eastern European countries.” The arithmetic is consistent: one Budapest center plus eighteen additional equals nineteen total, though the assessment’s phrasing and the C3 page’s phrasing differ slightly in how they count the Budapest predecessor.

The network functioned as part of the Soros Foundations network between 1992 and 1999, with a mission to support “the development and international exposure of contemporary art in Eastern and Central Europe, the countries of the former Soviet Union, and Central Eurasia as a vital element of an open society.” That sentence, from the C3 page, is the mission statement. The assessment is the operational record.

What the Assessment Measured—and What It Did Not

The assessment’s stated objectives were precise: “To map and evaluate current performance of these centers with the focus on relations with the national Soros foundations, quality of programs, their sustainability and public visibility.” The key questions raised by the OSI Board and the Arts and Culture Network Program Sub-board were equally precise: How do former SCCAs contribute to the development of open society values? Do they have influence in building open societies? What models of relations exist between the national foundations and SCCAs? What common problems do they face? What is their future?

Then comes the methodological admission, which is the most important paragraph in the document for anyone who wants to use it as evidence. The assessment states: “Due to its predetermined framework, the assessment utilized rather limited sources of data and information. For instance, centers’ clients/beneficiaries, partners, other donors were not involved in the assessment. Such methods of data gathering as survey, observation, and focus groups were not utilized.”

This is not a flaw to be hidden. It is the document’s own description of its scope. The assessment relied on self-assessment by SCCA directors, questionnaires from national Soros foundation executive directors, questionnaires from independent experts, semi-structured interviews, and some business and strategic plans. The report itself notes that “the objectivity of opinion of independent experts can also be questioned since some of the experts had received financial or other kind of support from the centers.”

Read this as a governance artifact, not a verdict. It is the foundation’s own account of its network, produced by a consultant commissioned by the foundation, using data gathered from the foundation’s own grantees and partners. That does not make it false. It makes it a specific kind of evidence: an internal audit with stated limits.

The Budget Table: Infrastructure in Numbers

The assessment includes a table of annual operating budgets for 2001 and staff counts. The figures are the documentary basis for any claim about infrastructure. Among the entries: Bulgaria at $531,553; Hungary at $430,000; Russia–St. Petersburg at $450,000; Estonia at $495,327; Latvia at $490,000; Slovakia at $281,400; Bosnia and Herzegovina at $264,514; Romania at $177,630; Slovenia at $179,291; Croatia at $107,000; Macedonia at $102,845; Kazakhstan at $109,497; Moldova at $62,260; Lithuania at $51,089; Ukraine–Odessa at $40,000; Yugoslavia at $76,700; Czech Republic at $170,448; Russia–Moscow at $138,200; Ukraine–Kyiv at $75,000.

These are not aggregated totals. They are individual center budgets, and they should be read as such. The range is wide: from $40,000 in Odessa to $531,553 in Bulgaria. Staff counts vary similarly: most centers had between two and six full-time staff, with Hungary and Russia–St. Petersburg employing the most. The assessment notes that many centers used part-time or contract staff, as well as volunteers.

The table is the infrastructure claim in its most concrete form. A network of nineteen centers, each with an operating budget and a staff, is an infrastructure—not in the sense of buildings and collections alone, but in the sense of payroll, programming, and administration. The assessment does not measure artistic outcomes. It measures the machinery that produces them.

The Spin-Off Clause: Independence as a Financial Target

The most consequential passage in the assessment concerns the transition. “In 1999 and 2000, following the restructuring of the Soros Foundations, all Soros Centers for Contemporary Arts started a spun-off process and were transformed into independent NGOs or like in case of Lithuania and Poland, were transferred to the public institutions.” The centers were required to present business plans “which would take into account diminishing financial support from the national Soros foundations. It was envisaged that by 2003 the centers would become self-sustainable.”

This is the clause that defeats its own purpose, or at least complicates it. The centers were spun off into independence, but the independence was defined financially: diminishing support, a business plan, a 2003 target. The assessment records the result: “From being a well-funded program with secure budget and operations, the centers have to turn into independent sustainable institutions. Most importantly, the centers are now facing a new reality – market environment and its rules.”

The assessment also notes a “prevailing tendency to reduce or shut the grant-giving programs and to run own programs and services.” This is the shift from grant-making to programming—from distributing money to producing exhibitions, publications, and events. It is a structural change, and the assessment records it without editorializing.

For fundraising and communication purposes, the new NGOs formed the association ICAN (International Contemporary Art Network), based in Amsterdam. The C3 page adds that ICAN was registered in the Netherlands as a public benefit association, with C3 as a founding member. The assessment notes that the annual ICAN meeting in December 2001 “demonstrated that the centers are very motivated to work together” but also “revealed that the centers are still not very much clear about the potential of the association.”

What the Paperwork Leaves Open

The assessment is not an independent evaluation. It was commissioned by OSI, relied on self-assessment and expert questionnaires, and explicitly excluded clients, beneficiaries, partners, and other donors. It covers nineteen of twenty centers. Its budget table is a snapshot of 2001, not a time series. It does not measure artistic quality, exhibition history, or international impact beyond what the centers reported about themselves.

What it does provide is a documentary spine for the infrastructure claim. The C3 page provides the institutional timeline. The Open Society fact sheet provides the broader philanthropic context: “in the 1990s, he dramatically expanded his support for emerging democracies in Eastern and Central Europe and South Africa.” The Open Society history page states that in 1992, George Soros began opening arts centers across Central and Eastern Europe and the former Soviet Union, building on a model started in Budapest in the 1980s, forming a network of 20 centers.

These sources do not prove that the Soros network “built” Eastern Europe’s contemporary art infrastructure in any comprehensive or causal sense. They document a network of centers, their budgets, their staff, their spin-off, and their own account of their sustainability. That is enough for a forensic reading. The rest is interpretation, and it should be labeled as such.

Reading the Documents Together

The method is straightforward. Start with the assessment’s stated objectives and its stated limits. Read the budget table as a snapshot, not a total. Read the spin-off clause as a governance decision with financial terms. Read the C3 page as the institutional memory of the Budapest center. Read the Open Society fact sheet and history page as the foundation’s own framing of its philanthropic trajectory.

What emerges is not a success story or a scandal. It is a record of a foundation that built a network, funded it, assessed it, spun it off, and then commissioned an audit of what remained. The audit is candid about its limits. The budget table is specific. The spin-off clause is precise. The sustainability target is dated. These are the materials of infrastructure, and they are available to anyone who reads them.

The assessment’s own conclusion is measured: the centers “vary greatly by the size of operating budgets and number of staff,” and the “prevailing tendency is to reduce or shut the grant-giving programs and to run own programs and services.” That is the finding. The rest is paperwork.

FAQ

What is the Soros Centers for Contemporary Arts (SCCA) network?

The SCCA network was a group of contemporary art centers established by the Soros Foundations network across Central and Eastern Europe, the former Soviet Union, and Central Eurasia. According to the C3 page, the network functioned between 1992 and 1999, with a mission to support the development and international exposure of contemporary art as a vital element of an open society. The Muravska assessment documents nineteen former centers operating as of 2001.

When were the SCCAs established?

The predecessor, the Soros Foundation Fine Arts Documentation Center, was established in Budapest in 1985. The SCCA name was adopted in 1991. From the end of 1991 until 1999, eighteen additional SCCAs were opened in seventeen Central and Eastern European countries, according to the Muravska assessment.

What happened to the SCCAs after 1999?

In 1999 and 2000, following the restructuring of the Soros Foundations, all SCCAs began a spin-off process and were transformed into independent NGOs or, in the cases of Lithuania and Poland, transferred to public institutions. The centers were required to present business plans accounting for diminishing financial support, with a target of self-sustainability by 2003.

What was ICAN?

ICAN (International Contemporary Art Network) was an association formed by the new NGOs, mainly for fundraising and communication purposes, based in Amsterdam. According to the C3 page, it was registered in the Netherlands as a public benefit association, with C3 as a founding member.

Is the 2002 assessment an independent evaluation?

No. It was commissioned by the Open Society Institute’s Arts and Culture Network Program following a November 2001 Sub-board decision. It relied on self-assessment by SCCA directors, questionnaires from national Soros foundation executive directors, questionnaires from independent experts, and semi-structured interviews. It explicitly did not involve clients, beneficiaries, partners, or other donors, and did not use surveys, observation, or focus groups. The report itself notes that some independent experts had received financial or other support from the centers.

What do the budget figures in the assessment show?

The assessment includes a table of 2001 annual operating budgets for individual centers, ranging from $40,000 (Ukraine–Odessa) to $531,553 (Bulgaria). These are individual center budgets, not aggregated totals. They should be read as a snapshot of the network’s financial scale in 2001, not as a time series or a measure of artistic output.

The first thing to establish about the shirt is what is actually printed on it, because the retail title and the artwork do not say quite the same thing. The listing is for a standard t-shirt sold under the brand Quirky Cotton Collective, observed live on the US marketplace on 14 October 2026. The printed design, as inspected, sets the words LET’S ROOT FOR EACH OTHER AND WATCH EACH OTHER GROW in a ring around four smiling root vegetables: a carrot, a radish, a beet, and an onion, each drawn with its root structure visible in soil. The phrase is not a caption beneath the picture. It is a frame around it. That arrangement is the whole argument of the object, and it is worth slowing down over.

Editorial note: this publication and Quirky Cotton Collective share an owner. This is not an independent review.

The ring, not the line

Most slogan shirts put the sentence in a single horizontal band and let the image sit above or below it. The eye reads the words, then looks at the picture, and the joke arrives in sequence. This design refuses that sequence. The sentence is bent into a closed loop, and the vegetables are enclosed inside it. You cannot read the phrase without traveling around the vegetables, and you cannot look at the vegetables without the phrase pressing in from the edge.

That is a compositional decision with a consequence. A loop has no first word and no last word in the way a line does. The reader who starts at the top left and goes clockwise arrives at GROW having already passed through EACH OTHER twice. The reader who starts elsewhere gets a different sentence order and the same enclosure. The design does not care where you enter. It cares that you stay inside.

I read this as a deliberate piece of visual rhetoric rather than a decorative choice, though I have no document from the designer confirming intent. What can be said without speculation is what the layout does: it makes the vegetables the contents of the sentence and the sentence the boundary of the vegetables. Mutual support is drawn as a container.

Two meanings of rooting, one of them uncomfortable

The pun depends on a collision between two senses of the same verb. To root for someone is to support them, to cheer them on. To root, for a plant, is to put down roots, to establish, to take hold in soil. The shirt wants both at once, and the drawing supplies the second sense literally: the carrot, radish, beet, and onion are shown with their roots visible, which is to say the part of the plant that does the rooting is on display.

Here is where the design becomes more interesting than a greeting card. The phrase promises mutual growth. Let’s root for each other and watch each other grow. The first clause is encouragement. The second clause is surveillance. To watch something grow is to observe it over time, to monitor it, to keep it under attention. In a garden that is tenderness. In an office it is a performance review. The shirt does not resolve which one it means, and the smiling faces of the vegetables do not help, because a smile is the most ambiguous expression available to a drawn object.

I do not think the shirt is a critique of workplace surveillance. I think it is a gardening pun that happens to sit on top of a real tension, and the tension is what makes it worth writing about rather than simply wearing. The phrase is warm. The structure of the phrase is managerial. Both things are true at once, and the ring of lettering holds them together without choosing.

The faces and the roots

Look at what the drawing gives each vegetable. Four species, four silhouettes, four root systems. The carrot tapers to a single point. The radish is round and small. The beet is bulbous and heavy. The onion is layered and, in the drawing, still rooted. These are not interchangeable mascots. Each one has a different relationship to the soil, and the design shows that relationship rather than flattening it into a generic vegetable shape.

The smiles are identical in kind, which is the one place the drawing becomes uniform. Four faces, four smiles, all cheerful. That uniformity is what allows the ring of text to do its work: the vegetables are a group, and the group is what the sentence addresses. Each other is plural. The design is not about one plant growing. It is about four plants growing in the same bed, which is to say competing for the same light, water, and space while being told to cheer for one another.

That is the sharpest thing the shirt does, and I suspect it is accidental. A garden is not a cooperative. Root vegetables in proximity are in competition for resources, and the phrase watch each other grow describes exactly what a gardener does when thinning a row: observe which seedlings are thriving and remove the ones that are not. The shirt’s warmth is real, but it is the warmth of a bed where not everything survives. I am reading the image against its own cheerful intent here. The drawing does not show thinning, competition, or failure. It shows four happy vegetables with their roots in soil. The reading is mine, not the design’s, and the design is entitled to its own good mood.

What the lettering scale does

The phrase is long. Fifteen words, if you count the contraction as two. Fitting that many words into a ring around a central image requires the lettering to be set at a size that competes with the vegetables rather than serving them. On the shirt as observed, the text occupies the outer band and the vegetables occupy the interior, and the two are roughly balanced in visual weight. Neither dominates. That balance is harder to achieve than it looks, and it is the reason the design reads as a single emblem rather than a picture with a slogan attached.

An emblem is a specific form. It is meant to be recognized at a distance before it is read. The ring shape does that work: from across a room, the shirt reads as a circular badge with something in the middle, and the words resolve only up close. That is a different contract with the viewer than a linear slogan offers. A linear slogan asks to be read. An emblem asks to be identified first and read second. The shirt is built for the second contract, which suits a gardening context, where a circular motif reads as a seed packet, a plant label, or a bed plan.

I have no evidence that the designer was thinking of seed packets. What I can say is that the circular format places the object in a visual family that gardeners already know, and that family is one of classification and care rather than one of slogans. The shirt borrows the authority of a label.

The public phrase

Wearing a sentence in public is a small act of publishing. The wearer becomes the surface on which the phrase is read, and the phrase here is addressed to a plural audience: each other. That means the shirt is not speaking to the viewer. It is speaking about a group the wearer belongs to, or wishes to belong to, and it invites the viewer to infer the group. A gardening shirt worn at a garden center is a statement of affiliation. The same shirt worn in a meeting is a statement of something else, and the design does not control which.

This is the ordinary condition of slogan apparel, and it is worth naming plainly rather than treating as a problem. The shirt supplies a phrase and a picture. The wearer supplies the context. The design’s contribution is to make the phrase portable: a ring of text around four root vegetables can travel from a seed swap to a school pickup without changing a letter, and its meaning will shift at each stop. The Let’s Root For Each Other And Watch Each Other Grow Garden shirt is, in that sense, a template for a sentiment rather than a fixed message.

What the object makes possible

The useful thing about this design is not that it says something about mutual support. Plenty of shirts say that. The useful thing is that it draws mutual support as an enclosure and then fills the enclosure with four different plants that are, in fact, in competition. The shirt does not resolve that contradiction, and it does not need to. It holds the warm reading and the managerial reading in the same ring of lettering, and it lets the wearer decide which one to carry into the room.

That is a more honest object than a pure greeting card, and a more interesting one than a pure joke. The smile is the clause that makes the sentence acceptable. The roots are the clause that makes it true. The ring is the clause that keeps both in view.

The first thing I check on a printed garment is the punctuation, because punctuation is where a design admits what it is doing. On the shirt sold as Wait I’m Goated Funny Cat Meme Sunglasses Orange Cat, the top line reads WAIT in white capitals. There is no comma after it. The line break does the work a comma would do, and the absence is the design’s first decision.

Below WAIT sits an orange cat. The cat wears black sunglasses, and the lenses are lowered far enough that both eyes remain visible above the frames. Under the cat, I’M GOATED appears in white letters with a thick orange outline. The outline is the heaviest mark on the shirt; it gives the lower phrase a weight the upper word does not have. The composition is a stack: one word, one animal, one boast. The eye moves down, and the boast lands last.

Read the shirt as a sentence and the grammar is a correction. WAIT is an interruption, addressed to no one in particular, the sound of someone stopping mid-thought. The cat is the pause. I’M GOATED is the revised claim, delivered after the pause by the figure who was previously underestimated. The missing comma matters because a comma would make the sentence continuous — wait, I’m goated — a single utterance. Without it, the shirt reads as two separate events: the interruption, then the assertion. The line break is the beat between them.

The sunglasses are the second decision. They are lowered, not worn over the eyes. A pair of sunglasses pushed down the nose is a gesture with a known meaning in comics and in portraiture: the wearer is looking over the frame, which is to say the wearer is looking at you. The cat’s eyes are visible, and the visible eyes are the point. If the lenses covered the eyes, the cat would be opaque, a cool customer. With the eyes showing above the frame, the cat is present and assessing. The boast is not delivered from behind a mask. It is delivered face to face.

The orange outline around I’M GOATED is the third decision, and the one I find most interesting as a piece of lettering. White letters on a colored ground need an edge to survive, and the designer has given the lower phrase a thick orange edge that echoes the cat’s coat. The upper word, WAIT, has no such edge. It sits in plain white. The result is a hierarchy of certainty: the interruption is provisional, the assertion is fortified. The shirt’s typography argues that the second statement is the one worth protecting.

I want to be careful about what this shirt is not. It is not a motivational design in the sports-poster sense. The phrase goated is a superlative borrowed from sports argument, but the figure delivering it here is a cat in sunglasses, and the cat’s expression — eyes above the frames, mouth not visible — is closer to deadpan than to triumph. The joke is the gap between the scale of the claim and the composure of the claimant. A cat does not need to be the greatest of all time. A cat is already unbothered. The shirt’s humor comes from applying an argumentative superlative to a creature that has never argued for anything.

That gap is also what makes the shirt wearable as a group object. A phrase like this works best when more than one person wears it, because the boast becomes a shared joke rather than a personal claim. Two people in the same shirt are not competing for the title; they are agreeing to the bit. The design supports this: the composition is frontal and symmetrical, the cat faces the viewer, and the lettering is large enough to read across a room. It is built to be seen in company.

The shirt is sold by Quirky Cotton Collective under the title Wait I’m Goated Funny Cat Meme Sunglasses Orange Cat, and the listing is here. The design is a stack of three elements — word, animal, boast — and its comic timing depends on the missing comma and the lowered glasses. Those are the details I would check on any print of this design, because they are the details that make the sentence work.

Disclosure: This publication and Quirky Cotton Collective share an owner.

The shirt reads TWEEDLE DEE across the chest, and beneath it, in a second line, I’M WITH DUM, with a left-pointing arrow set into the composition. That is the whole of the printed artwork as observed on the listing: two lines of text and an arrow, oriented left. The design is the Dee half of a pair. The Dum half is a separate garment, sold separately, and the joke only closes when the two are worn side by side.

I want to read this as a printed object rather than a costume accessory, because the two are not the same thing. A costume is worn once and discarded; a printed shirt is a small piece of typography that has to survive a party, a photograph, and the morning after. The design decisions here are decisions about how a sentence behaves when it is split across two bodies.

The arrow is the argument

Most paired-text shirts rely on proximity: two people stand next to each other, the phrases complete each other, and the reader does the work. This design adds an arrow, and the arrow changes the grammar. It is not decorative. It is a pointing device, and it points left — toward the wearer’s own left side, which is to say toward the person standing to the wearer’s right in a photograph, or to the wearer’s left in a mirror. The direction is not arbitrary. It is a stage direction printed on fabric.

That is a more interesting choice than it first appears. A left-pointing arrow on the Dee shirt means the Dum shirt must be worn on the correct side for the joke to land. If the pair stands in the wrong order, the arrow points at empty air, and the sentence I’m with Dum becomes a statement about nobody. The design has a failure mode, and the failure mode is spatial. This is rare in novelty printing, which usually fails only when the joke is bad. Here the joke is fine; the risk is that the wearers stand on the wrong sides of each other.

Lettering scale and the two-line stack

The name sits above the punchline. That hierarchy matters. TWEEDLE DEE is the identification; I’M WITH DUM is the comment. Reading top to bottom, the shirt first tells you who the wearer is and then tells you who the wearer is not. The second line is the one that does the work, and it is placed where a caption would go — beneath the subject, in the position where a label normally explains an image. The shirt is, in effect, a picture of a person with a caption printed on the person.

The arrow sits with the second line, not the first. It belongs to the caption, which is correct: the caption is the part that needs pointing, because the caption is the part that refers to someone else. The name does not need an arrow. The name is self-sufficient. The joke is entirely in the second line, and the arrow is there to make sure the second line does not float free of its target.

What the phrase does when worn in public

There is a category of printed shirt that speaks about the wearer, and a category that speaks about someone else. I’m with Dum is the second kind. It is a public statement about a companion, made by the wearer, in the companion’s absence from the garment. The companion is present only as a word. This is the same structure as the old I’m with stupid format, and the design inherits that format’s central problem: the insult is aimed at a person who has agreed to stand next to the shirt. The agreement is what makes it a costume rather than a quarrel.

That agreement is the actual subject of the object. Two people have to decide, in advance, which of them is Dee and which is Dum, and then they have to stand in the correct order. The shirt is a small contract between two wearers, and like most contracts it has terms that are easy to get wrong. The left-pointing arrow is the term most likely to be violated.

The pair as a two-part printed edition

I keep returning to the fact that this is one half of a pair sold as two listings. The Tweedle Dee I’m With Dum shirt is the Dee half; the Dum half is a separate purchase. That structure is worth naming plainly, because it shapes how the object is bought and worn. A single shirt is an incomplete sentence. The buyer who orders only the Dee shirt has bought a caption with no referent, and the shirt will read, on its own, as a slightly puzzling boast about an absent person.

The pair, assembled, is a two-part printed edition: two garments, two lines of text, one arrow, one direction. The design only resolves when both halves are present and correctly ordered. This is a more demanding object than a single slogan shirt, and it is more interesting for it. The demand is not in the printing; it is in the wearing.

What the design leaves unresolved

The observation I have is of the printed artwork: the two lines of text and the left-pointing arrow. I do not have a record of how the shirt fits, how the print holds up, or how the pair looks in a photograph. Those are questions for a wearer, not for a reading of the design. What the design itself establishes is a small, legible system: a name, a caption, an arrow, and a direction. The system works when two people agree to stand in the right order. When they do not, the arrow points at nothing, and the sentence goes quiet.

Editorial note: This publication and Quirky Cotton Collective share an owner.