Food writing that goes beyond the recipe.
Food is culture, memory, science, and pleasure. We write about all of it. Recipes that actually work because we test them until they do. Restaurant reviews that treat the craft with respect. Stories that dig into where ingredients come from and why cooking techniques matter.
Topics we cover: Recipes · Restaurant Reviews · Technique · Ingredients · Food Culture · Travel
The garment preview shows two lines of white type on a dark ground. The first line reads IN DOG BEERS. The second reads I’VE ONLY HAD ONE. There is no illustration — no dog, no glass, no leash. The joke is carried entirely by the substitution of one word and the line break that delays it. This is a typographic pun, and the shirt is the page it is printed on.
That restraint is worth pausing over. A lesser design would have added a cartoon beagle with a pint. This one trusts the reader to hear the echo of the familiar phrase — in dog years, I’ve only had one — and to feel the small click when years becomes beers. The white letters sit close together, the first line slightly wider than the second, so the eye lands on DOG BEERS before it resolves the sentence. The delay is the joke.
What the shirt does, then, is not tell a drinking story. It stages a unit conversion. Dog years are a folk arithmetic: one human year equals seven dog years, or so the old rule goes. The shirt borrows that arithmetic and applies it to beer. If one beer is seven dog beers, then the wearer who has had a single beer has, in dog-beer terms, had seven. The punchline — I’ve only had one — is a plea of innocence that the first line has already undermined. The math is the mischief.
I am not claiming the shirt invented that joke, nor that any particular algorithm promotes it. I am reading the object as it appears: a dark shirt, white sans-serif capitals, a two-line setup and a two-line fall. The design is a piece of typographic timing, and its humor depends on the reader’s familiarity with a phrase that belongs to pet ownership and barroom banter alike. The shirt does not explain itself. It assumes you have heard the original.
That assumption is what makes it a useful object for this column. Cooks and drinkers spend their working lives reading labels, tickets, and menus — short texts that must land in a hurry. A shirt like this is a label turned inside out: a piece of language that expects to be decoded at a glance, then rewards a second look. The dog beers shirt is not a costume. It is a sentence with a seam in it.
The dark ground matters. White type on a dark field is the grammar of the chalkboard, the bar menu, the specials card. It reads as provisional, handwritten even when it is set in a clean face. The shirt borrows that grammar without pretending to be a chalkboard. The letters are too even for that. But the contrast does the work of a sign: it asks to be read across a room, not studied up close. The joke is built for the distance between two barstools.
There is a small editorial note to make here. This publication and Quirky Cotton Collective share an owner. This is not an independent review. That fact does not change the reading of the type, but it belongs in the record, the way a menu should say who owns the kitchen.
What the shirt does not do is tell us anything about the person wearing it. It does not prove they drink, or own a dog, or have a sense of humor that survives a second reading. It is a surface with a sentence on it. The sentence is the object. The rest is inference, and inference is cheap. The design is what we can actually see: two lines, one substitution, a dark field, white letters. The joke is old enough to be a folk form, and the shirt is a clean copy of it. That is enough.
Return to the preview. IN DOG BEERS above I’VE ONLY HAD ONE. The first line is the setup, the second is the alibi. The shirt is a small machine for making the alibi fail. It does not need a dog, a glass, or a story. It needs a reader who knows the phrase and a room dark enough for white letters to carry. That is the whole design, and it is complete.
In January 2024, Thailand’s Government Public Relations Department published a short news item under the headline “Government Boosts Global Thai Cuisine Appeal.” It is not a long document. It announces a Ministry of Commerce campaign called “Travel, Taste, Eat Thai SELECT,” led by the Department of International Trade Promotion (DITP), and it contains the numbers that most coverage of Thai food diplomacy repeats: government support since 2002, restaurant counts abroad rising from 5,500 to 15,000, and a current global total of 17,478 Thai restaurants, of which 1,546 hold Thai SELECT certification, “mainly concentrated in the US” (Government Public Relations Department, 10 January 2024).
Those figures are usually read as a branding success story. This piece reads them differently, starting from the one part of the campaign that functions as a document rather than a slogan: the certification checklist. A checklist that specifies who works in the kitchen and where the ingredients come from is not only a food-authenticity instrument. It is a labor instrument wearing a menu’s clothing.
The most detailed description of Thai SELECT criteria in the retrieved record comes from a secondary summary published by SEAMEO RECFON, which describes the program as requiring that a restaurant be open at least five days a week, employ government-trained Thai chefs, and use genuine Thai products (SEAMEO RECFON, “The Gastrodiplomacy Series – Part 1”). The same summary attributes the certification to the Thai government and describes it as intended to ensure quality and promote authenticity.
Note what is and is not on that list. There is no recipe requirement. There is no fixed roster of dishes, no mandated pad Thai formula, no specification for wok heat or tamarind concentration. The criteria are about operating days, staffing, and sourcing. In other words, the document polices the conditions of production, not the plate. That is why it belongs in a labor history as much as a culinary one.
One caution belongs here. The SEAMEO summary is a secondary source, and the official Thai SELECT criteria document itself was not retrieved for this piece. The wording above should be treated as the summary’s description of the criteria, not as a verbatim quotation from DITP. The distinction matters if you intend to rely on the checklist in your own writing: get the application form or inspection sheet from DITP before treating any clause as settled text.
The government release gives a clean before-and-after: support since 2002 credited with moving overseas Thai restaurants from 5,500 to 15,000, and a present count of 17,478, of which 1,546 are certified. The SEAMEO summary repeats the 17,478 and 1,546 figures and adds that the US leads with roughly 6,850 restaurants, about 39 percent of the total abroad.
These are restaurant counts. They are not employment records. Nowhere in the retrieved material is there a figure for how many cooks, dishwashers, servers, or managers work in those establishments, how many are Thai nationals, how many arrived under what status, or what they are paid. The campaign’s public accounting measures the spread of a brand, not the condition of the people who execute it. That is not a scandal in itself; trade promotion agencies count businesses. But it means the campaign’s own success reporting cannot answer the labor question, and readers should not let a rising restaurant count stand in for evidence about kitchen work.
The SEAMEO summary separates two initiatives that are frequently conflated. Global Thai, which the summary says began in 2002 and gave the term gastrodiplomacy its early currency, is described as emphasizing the internationalization and spread of Thai cuisine, and as working through financial incentives for Thai nationals to open restaurants abroad. Thailand: Kitchen of the World is described as highlighting culinary heritage, recipes, and food products, and as providing the foundation for Global Thai. Thai SELECT is presented as Global Thai’s certification component.
That division of labor is worth keeping straight. One campaign moves restaurants and people; the other moves ingredients and reputation. The certification sits at the junction, because its sourcing clause ties a restaurant’s authenticity claim to Thai product exports while its staffing clause ties it to Thai labor.
Of the three criteria in the summary, “employ government-trained Thai chefs” carries the most weight and receives the least documentation. It implies a training pipeline, a credential, and a labor market in which that credential has value. It also implies a boundary: restaurants that employ cooks trained elsewhere, or cooks of other nationalities, sit outside the certified category regardless of what they serve.
The retrieved sources do not answer the questions this clause raises. Who conducts the training, and where? Who pays for it? Does the credential travel with the cook across borders, and under what immigration status does a certified chef work abroad? What happens to the uncertified cooks in the same city, many of them immigrants running small kitchens that will never appear in a DITP count? These are not rhetorical questions. They are the evidence a labor reading of the campaign requires, and they are absent from the record assembled here.
It is worth being explicit about the limits. No retrieved source gives a figure for Thai chefs trained, migrated, or employed under the campaign. No source documents visa categories, wages, or working conditions for Thai chefs abroad. Claims of that kind should not be made from this material, and this article does not make them.
The most interesting number in the government release is the ratio. Roughly nine in ten Thai restaurants worldwide are not Thai SELECT certified. Some of that gap is administrative: certification is voluntary, costs time, and may not be worth it for a small operator. Some of it is geographic, since certification clusters in the US. But whatever its causes, the gap describes a two-tier landscape.
Consider a hypothetical pair, clearly labeled as illustration rather than reportage: two restaurants in the same city, both serving pad Thai, both run by Thai-born owners. One pursues certification, hires credentialed chefs, and sources Thai products through channels the program recognizes. The other is a family operation with two cooks, no certification, and a supplier list that includes local substitutes. A diner looking for “authentic” Thai food may read the Thai SELECT logo as a signal and choose the first. The second restaurant’s labor, often the more precarious and the more typical of immigrant food businesses, remains uncounted, unsupported, and unmarked. The certification does not create that hierarchy by itself, but it formalizes it, and the formal version is the one that appears in the statistics.
Pad Thai appears in the SEAMEO summary as Thailand’s national dish and a global ambassador for Thai cuisine, elevated as part of the campaign. That is a description of the dish’s promotional role, not a history of its origins. The retrieved sources do not document pad Thai’s wartime promotion, its recipe lineage, or the process by which it acquired national-dish status, and this article does not supply those claims from memory. If the dish’s biography matters to your argument, it needs its own primary documents: cookbooks, government publications, or menu records from the relevant decades.
The same restraint applies to the campaign’s downstream effects. The SEAMEO summary notes that Thai tourism rose from roughly 10 million visitors in 2001 to 39.8 million in 2019 and asserts that Global Thai contributed to that growth. A correlation of that kind, stated in a secondary source without a causal design, cannot bear the weight of a causal claim. Restaurant counts, tourist arrivals, and export values moved together over two decades in which many other things moved too.
If the goal is to understand what the restaurant-seeding campaign did to the people who cook, the document trail to assemble is specific. First, the official Thai SELECT criteria and application form from DITP, to replace summary language with the actual clauses. Second, the original Global Thai program document from 2002, stating objectives and budget, which the retrieved set lacks. Third, any Ministry of Labour overseas placement statistics for chefs and food-service workers in the campaign’s target countries. Fourth, inspection or certification records for individual restaurants, which would show which establishments were approved, when, and on what evidence.
Until those documents are in hand, the honest reading of the January 2024 release is narrower than the headlines suggest. It records a government counting restaurants and certifying a minority of them against criteria that include staffing and sourcing. It does not record what happened inside those kitchens. The checklist is the place to start, because it is the one artifact in the campaign that names labor at all, and naming labor is the first step toward counting it.
The retrieved sources describe Global Thai as providing financial incentives for Thai nationals to open restaurants in foreign countries, per the SEAMEO summary. They do not document direct government ownership or operation of restaurants. The precise incentive mechanisms are not described in the retrieved material.
According to the January 2024 government release, 1,546 of 17,478 Thai restaurants worldwide held certification, with the certified group concentrated mainly in the US.
The secondary summary describes criteria covering operating days, government-trained Thai chefs, and genuine Thai products. It does not describe recipe requirements. The official criteria document was not retrieved, so this should be confirmed against DITP’s own materials.
The SEAMEO summary describes it that way in the context of the campaign. The retrieved sources do not document the formal designation or its history, so treat the label as the campaign’s framing rather than an established administrative fact.
The shirt does not show a garden. It shows a cross-section. Four root vegetables — carrot, radish, beet, and onion — sit in a band of soil, each with a face, each with its edible root drawn downward into the dirt. Above them, in capital letters, runs the line: LET’S ROOT FOR EACH OTHER AND WATCH EACH OTHER GROW. The lettering wraps the group rather than sitting above it, so the words and the vegetables occupy the same frame. Nothing else is in the picture. No hands, no trowel, no fence, no sky.
That framing is the first thing worth noticing, because it makes an argument the sentence alone does not. A gardening shirt could have shown a harvest: baskets, a proud row of seedlings, a sun. This one shows the part of the plant that stays underground, the part you are not supposed to see until you pull it. The roots are visible, labeled by their own shapes, and given faces. The design asks you to look at the labor that usually happens out of sight and treat it as the point.
The pun works on two senses of root. To root for someone is to support them from a distance, often without any power to change the outcome. To put down roots is to stay, to anchor, to grow in place. The shirt uses both at once: the vegetables are literally rooted in soil, and the sentence asks the reader to be the kind of person who roots for others. The second line, and watch each other grow, completes the thought. Support is not passive here. It involves attention — watching — and it expects change.
I read the four vegetables as a small ensemble rather than a set of mascots. Carrot, radish, beet, and onion are different species with different growing habits, different colors, different kitchen uses. The carrot is orange and tapered; the radish is small and red; the beet is round and dark; the onion is layered and pale. They are drawn together in one soil band, smiling, roots exposed. The design does not rank them. It does not put the carrot in front. That evenness is doing some of the work of the sentence: mutual support among things that are not the same.
There is a tension in the phrase watch each other grow that I keep returning to. Growth is not always comfortable to watch. In a garden, growth is uneven; some plants crowd others out, some bolt, some fail. In people, the same is true. A shirt that says watch each other grow is making a generous claim — that another person’s growth is something to welcome rather than manage. It is a nicer sentiment than most slogan shirts attempt, and the artwork supports it by keeping all four faces level and smiling. No one is looking up at anyone else.
The retail listing describes the design as a gardening and mutual-growth pun, and the visual observation confirms the four smiling root vegetables with their roots visible in soil. That is the object. I have not worn it, and I am not going to pretend otherwise. What can be read from the artwork is the arrangement: the words surround the vegetables, the soil line cuts across the middle, the faces are simple and open. The palette in the observed design is the natural one — orange, red, deep purple-red, off-white — set against the soil. It reads as a garden cross-section more than a cartoon.
For cooks, the shirt has a second life. Carrot, radish, beet, and onion are all things you pull, scrub, trim, and cook. They are the vegetables that teach you what a root actually is: a storage organ, a way for a plant to survive a season and come back. A beet pulled from the ground is a small lesson in patience. An onion is a lesson in layers. The shirt does not say any of this, but the choice of vegetables makes it available. It is a design that rewards a second look in the kitchen, not just in the garden.
The phrase let’s root for each other also carries a quiet correction. It does not say let’s compete or let’s compare growth. It says root, which in the gardening sense means to establish, and in the supportive sense means to cheer. The double meaning is the whole joke, and it is a gentle one. The shirt is not making a demand. It is proposing a posture. That is a different thing from a motivational slogan, and it is why the design feels warmer than its genre usually allows.
If you are looking for the lets root for each other shirt, the listing identifies it as a standard t-shirt from Quirky Cotton Collective, with the artwork described above. I would read the design before reading the product copy. The design is the argument: four different roots, all visible, all smiling, all growing in the same soil. The sentence is the caption. The shirt is the cross-section.
Editorial note: This publication and Quirky Cotton Collective share an owner.
Begin with one document and stay with it. The document here is a 2020 Curious City feature by Monica Eng for WBEZ, which reconstructs Chicago restaurant life from roughly 1885 to 1920 using archival menus, newspaper clippings, and interviews with food historians. It is a secondary source, not a primary one, and that matters for what follows. But it is dated, it is specific, and it contains a passage that functions like a ledger entry: a Clark Street saloon owner, quoted through the Tribune in 1913, explaining why he closed.
“Some of those hobos spend a nickel for a glass of beer and eat 20 cents worth of liver,” he said. “Do you get me? That’s why I had to close up.”
That sentence is the whole argument in miniature. The free lunch was not charity and it was not hospitality in any sentimental sense. It was a subsidy paid in food, priced against a drink, and withdrawn at the point where the subsidy exceeded the margin. The owner’s complaint is not about gluttony. It is about arithmetic.
Eng’s piece states that early twentieth-century Chicago had about 8,000 saloons, and that to stay competitive they began offering free lunch with a drink — a practice the article describes as controversial with temperance advocates. The examples are concrete. Historian Danny Block is quoted describing a line of bars outside the stockyards where a drink came with a free hot oyster for lunch. In the Loop, First Ward Alderman Hinky Dink Kenna gave beer drinkers hot soup and unlimited bread at the Workingman’s Exchange on Clark. And the Berghoff, still operating, gave away corned beef sandwiches; a spokesperson told Eng that Herman Berghoff, as a recent German immigrant, had trouble getting a plain liquor license and could only obtain a retail license that required him to serve food.
Read those four examples together and a system appears. The free lunch was a competitive instrument, a licensing workaround, and a caloric subsidy, all at once. It was also, by the document’s own account, a male institution. Eng writes that respectable women were not seen in — or often allowed in — bars at the time, so the cheap saloon lunch was “pretty much only offered to men.”
The nickel-beer, twenty-cent-liver complaint is worth taking literally. It describes a transaction in which the customer’s outlay is fixed and small, and the house’s cost is variable and controlled by the customer’s appetite. That is an unstable arrangement. Eng reports that in 1913 two saloons went broke after giving away too much liver and onions. Whether or not the Tribune account is exact in every detail, the structure it records is clear: the free lunch had a ceiling, and the ceiling was set by what a saloon could recover on drink sales.
This is where the complaint justifies a measure of indignation, and where it also limits it. The owner’s words frame the customer as a problem — “those hobos” — but the same words record a man who ate twenty cents’ worth of food on a nickel. That is not extravagance. That is a working person extracting the maximum from a system designed to extract from him. The document does not tell us who these customers were beyond the owner’s label, and it does not tell us what they earned. It tells us what they could get for a nickel, and what the house could stand to lose.
The article calls the free lunch “controversial with temperance advocates.” That is the document’s assertion, and it is the only thing the document gives us on the subject. It does not quote a temperance pamphlet, a sermon, an organization’s resolution, or a dated statement of opposition. It does not explain the temperance argument against the table.
This is an evidentiary gap, and it should be named rather than papered over. A reader who wants to know why temperance advocates objected to the free lunch — whether they saw it as a lure, a subsidy of drinking, a degradation of the meal, or something else — will not find the answer in this source. Reconstructing that argument from memory or inference would violate the method. The honest statement is: the document records that the practice was controversial with temperance advocates, and it does not record their reasoning. A dated temperance text would be the next document to read.
The more consequential exclusion the document records is not class but gender. The same cheap calories were available to male drinkers and structurally unavailable to the shop girls and office workers Eng describes flooding the Loop as typists and clerks. Eng notes that respectable young women could not take part in the cheap saloon lunches, and that progressive women’s groups stepped in to fill the need — opening low-cost lunchrooms, including one near Jane Addams Hull House, and “suffrage restaurants” in the Loop.
That parallel table is the more important labor story, and the document supports it more thinly than it supports the saloon. We know the lunchrooms existed and roughly who ran them and why. We do not have their menus, their prices, or their payrolls in this source. What we can say is that the saloon’s free lunch was not a universal working-class institution. It was a gendered one, and the women’s lunchroom was the correction — not a charitable afterthought but a separate food system built because the first one was closed to half the workforce.
Eng traces the lunchrooms giving way to larger cheap lunch joints called cafeterias, which kept prices low by eliminating servers and tips. Thompson’s Cafeteria had more than a dozen Loop locations by 1910 and, to keep customers moving, sat diners in one-armed schoolroom desks that didn’t encourage lingering.
The document gives us the facts; the reading that follows is interpretation, and should be labeled as such. Removing the server removes a wage relationship — the tip disappears, and with it a category of worker whose earnings the document does not record. The one-armed desk is furniture that enforces throughput. Neither detail is presented in the source as a labor critique; both are presented as cost control. But cost control is a labor decision, and the cafeteria’s efficiency was achieved partly by deciding which workers would not be employed. A payroll record for cafeteria staff would let us say more. The document does not supply one.
This piece has stayed close to a single secondary source on purpose. From it we can say that Chicago saloons in the early twentieth century offered free lunch with a drink as a competitive practice; that the practice had specific local forms — the stockyard oyster, Kenna’s soup and bread, the Berghoff’s licensing-required corned beef; that it could be eaten out of existence, as the 1913 closures and the owner’s nickel-and-liver arithmetic show; that it was gendered; and that women’s lunchrooms and cafeterias developed alongside it.
We cannot say, from this document, what temperance advocates actually argued, how the free lunch worked outside Chicago, whether Prohibition ended it, or what the workers who cooked and served it were paid. Those are not gaps to be filled with plausible-sounding prose. They are the next documents to find: a dated saloon menu or license showing the free-lunch terms, a temperance pamphlet stating the objection, a women’s lunchroom menu, and a payroll record for the kitchen and counter staff. Until those are in hand, the complaint stands as it is — a nickel against twenty cents’ worth of liver, and a man explaining why he closed.
No. The document records it as free with a drink. The customer paid for the beer or whiskey; the food was the house’s cost, and the 1913 closures show that cost could exceed what drink sales recovered.
The WBEZ article describes the practice as “controversial with temperance advocates.” It does not quote or summarize their argument. That claim is recorded here as the document’s assertion, not as an established finding about temperance reasoning.
According to the document, respectable women were not seen in or often allowed in bars at the time, so the cheap saloon lunch was pretty much only offered to men. Women’s low-cost lunchrooms and suffrage restaurants in the Loop are described as filling that need.
The document traces low-cost lunchrooms giving way to cafeterias such as Thompson’s, which eliminated servers and tips and used one-armed desks to keep customers moving. It does not claim that cafeterias replaced the saloon lunch directly, and neither does this article.
The graphic is a butter stick turned upright and worn: a pale yellow block, its wrapper rendered as a flat printed panel, with the words SALTED BUTTER set in blue capitals that run vertically down the length of the stick. The lettering is the kind of plain, condensed sans-serif you find on a supermarket dairy case, not a hand-lettered farm label. The blue is the blue of a printed wrapper, slightly cool against the yellow. What the artwork does, before any joke lands, is reproduce the grammar of packaging: a product name, a color code, a shape that reads as a unit of sale. The shirt does not depict butter. It depicts the label on butter, and then stretches that label to the size of a torso.
That distinction is the whole argument. A costume that showed a stick of butter would be a picture of food. This one is a picture of a package, and the package is doing the comic work. The design says salted, not unsalted — the observation is literal, the blue capitals read SALTED BUTTER and nothing else — and that single word is the hinge. Salted butter is the default butter, the one most American cooks reach for without thinking, the one that comes in the familiar quarter-pound stick with the familiar blue-and-yellow wrapper. Unsalted is the fussier sibling, the one recipes specify when they want control over seasoning. By printing SALTED and stopping there, the shirt takes the unmarked, everyday choice and makes it a declaration. It is a costume about being the default.
I read the vertical lettering as a deliberate formal choice rather than a decorative one. On an actual butter stick, the wrapper text runs along the length of the stick because that is how the stick is oriented in the box and in the hand. Turning the text vertical on a shirt means the wearer’s body becomes the stick: the letters run from shoulder to hem the way they would run from end to end on the product. The costume only works if the proportions are roughly right, and the graphic commits to that proportion. It is not a logo placed on a chest. It is a wrapper wrapped around a person.
The retail listing frames the object as a couples costume, and the brand name on the listing — Salted and Unsalted Butter Couples Costume — makes the pairing explicit. That framing is worth pausing on, because it reveals what the single shirt cannot say by itself. One shirt says SALTED. The joke only completes when someone else stands nearby saying UNSALTED, or when the wearer is understood as half of a pair. The design is therefore a fragment of a larger bit: the two butter types as a matched set, the default and the deliberate, the everyday and the recipe-specific. A salted butter shirt worn alone is a label without its counterpart, which is either the point or the setup, depending on the room.
What the shirt does not do is make a nutritional claim, endorse a brand, or specify a fabric. The artwork is a printed graphic, and the observation that matters is visual: blue vertical capitals, a yellow stick, the word salted. Everything beyond that — how the cotton feels, how the print holds up, whether the fit runs true — is not in the record I have, and I will not invent it. The design can be read from the image alone. The wearing cannot.
There is a small culinary irony in the choice of word. In most professional kitchens, unsalted butter is the working butter, because it lets the cook control the salt. Salted butter is what goes on the table, what gets spread on bread, what sits in the dish at the diner. So a shirt that announces SALTED is, in a quiet way, a shirt about the dining room rather than the line. It is the butter of the toast plate, not the sauté pan. That reading is my interpretation of the word’s associations, not a claim about the designer’s intent, but it fits the costume logic: this is butter as it appears to the person eating, not butter as it appears to the person cooking.
The broader genre here is food packaging worn as clothing — the costume that turns a grocery item into an identity. The butter stick works because the packaging is already iconic and already legible at a distance. The blue capitals are the kind of type that survives being printed on fabric and read across a room. The joke does not require a caption or a speech bubble. It requires only that the viewer recognize the wrapper, which most people who have ever opened a refrigerator will.
If you are considering the salted butter shirt as a Halloween piece, the useful question is not whether it is funny in the abstract but whether the pairing is available to you. The design is built to be half of a set. Worn alone, it is a label. Worn with its unsalted counterpart, it becomes a sentence. That is a structural observation about the artwork, not a review of the garment, and it is the thing I would want to know before ordering: the shirt is a punchline that expects a second person.
Return, then, to the blue letters. They are plain, vertical, and specific. They say salted, not unsalted, and in that refusal to be the neutral option the costume finds its small comic edge. The wrapper is the joke, and the joke is that the default has decided to announce itself.
Editorial note: This publication and Quirky Cotton Collective share an owner.
The shirt does not show a pumpkin. It shows a pumpkin, a spoon, and a farmer in a straw hat, arranged beneath four words set in lettering that wants to be read aloud: PETER PETER PUMPKIN EATER. The farmer stands to one side, the spoon is held upright, and the pumpkin sits at the center like a prop waiting for its scene. This is not a photograph of a costume. It is a diagram of a rhyme, and the rhyme is the point.
I have been looking at the listing image for this Peter Peter Pumpkin Eater couples costume shirt the way I look at a menu from a kitchen I will never eat in: for what it chooses to show and what it leaves off. The design is spare. The lettering is the largest element, the farmer is the smallest, and the spoon is the hinge between them. That hierarchy matters. The shirt is not selling a pumpkin. It is selling a sentence that most English-speaking adults can finish without thinking. The pumpkin is just the punctuation.
The rhyme itself is older than the shirt and older than the holiday it is being sold for. It appears in early nineteenth-century collections of nursery rhymes, and its history is tangled enough that I will not pretend to trace it here. What matters for this object is that the rhyme has a structure: a man, a wife, a pumpkin shell, and a problem of containment. The shirt reduces that structure to three visual tokens and a name. The farmer in the straw hat is Peter. The spoon is the tool of the eater. The pumpkin is the shell. The wife is absent, which is either an oversight or the quiet joke of the design. I read it as the joke. A couples costume that shows only one half of the couple is making a claim about which half is the costume.
This is where the shirt becomes more interesting than a seasonal novelty. Halloween couples costumes usually work by pairing: salt and pepper, ketchup and mustard, a plug and an outlet. The joke is in the match. This shirt does not match anything on its own. It names Peter and leaves the wearer to supply the rest. The second shirt, if there is one, is the wife in the pumpkin shell, and the design gives no instructions for her. That absence is the design’s real content. It is a costume that requires a collaborator, and the collaborator is not pictured. The spoon is the only object that could belong to either of them.
I keep returning to the spoon. In the rhyme, Peter puts his wife in a pumpkin shell and keeps her very well. The spoon is not in the rhyme. It is an addition, and it is the addition that makes the shirt legible as a food object rather than a fairy tale. A spoon is a serving tool. It implies a bowl, a liquid, a meal. The pumpkin becomes a vessel, and Peter becomes a cook, or at least a man with a utensil. The shirt does not say he is a cook. It says he is an eater, and the spoon is there to remind you that eating is a kind of work. That is the reading I would defend: the shirt is not about a pumpkin. It is about the labor of feeding, and the farmer’s straw hat is the only clue that the labor happens outdoors, in a field, before the pumpkin ever reaches a table.
The lettering is set in a style that reads as storybook, slightly uneven, the kind of type that wants to look hand-drawn without committing to a hand. It is the same register as a children’s book cover, which is appropriate for a nursery rhyme and also a little evasive. A storybook font tells you not to take the image too seriously. It softens the rhyme’s darker implication, the wife in the shell, the containment. The farmer’s face is not visible in the observation I have. He is a hat and a posture. That blankness is useful. It lets the wearer project whatever Peter they need: a partner, a joke, a warning.
What the shirt does not do is resolve the rhyme’s problem. It does not show the wife, it does not show the shell as a shell, and it does not show the keeping. It shows the ingredients of the story and stops. That is the correct decision for a costume. A costume is an invitation to a role, not a summary of a plot. The wearer supplies the rest, and the second wearer supplies the wife, and the pumpkin on the shirt is the only thing both of them share. The spoon is the only thing either of them holds. I find that a more honest image of a couples costume than the matched pairs that line the seasonal aisles: two people, one object, and a rhyme that neither of them wrote.
The shirt is sold as a couples costume, and the listing title says so. The design itself does not. That gap between the retail description and the printed artwork is worth naming, because it is where most costume shirts live. The title promises a pair. The image delivers a single figure and a utensil. The buyer who reads the shirt literally will see a man in a hat. The buyer who reads the rhyme will see the whole story, including the part the shirt leaves out. Both readings are available, and the shirt does not choose between them. That is not a flaw. It is the mechanism.
I would not wear this shirt to a pumpkin patch. I would wear it to a table, or to a kitchen where someone is doing the work the spoon implies. The shirt is a costume for the person who feeds, and the pumpkin is the dish, and the farmer is the one who grew it. The rhyme calls him an eater, but the design calls him a grower, and the spoon is the tool that turns one into the other. That is the argument I would make for this object: it is not a pumpkin shirt. It is a shirt about the distance between a field and a bowl, and the person who closes it.
Editorial note: This publication and Quirky Cotton Collective share an owner.
The hat is cream, engraved-style, pointed, with a curved brim and a square buckle. Above the lettering it sits like a seal on a document. Below it, in serif capitals, the shirt reads YES I CAN DRIVE A STICK. A long broom runs down the right side of the composition, and a crescent moon, small stars, and flourishes surround the words. This is the artwork on the Quirky Cotton Collective shirt listed as Yes I Can Drive A Stick Funny Witch Broom Halloween Women, observed on the live US listing on October 14, 2026. The design is not a photograph or a slogan slapped on a blank. It is a small piece of visual rhetoric, and the rhetoric depends on a pun that only works if you see the broom before you read the sentence.
The phrase itself is ordinary enough. Drive a stick is colloquial for operating a manual transmission, and it carries a faint swagger — the implication that the speaker can handle a machine that demands coordination. Placed under a witch’s hat and beside a broom, the same words become a claim about broomstick piloting. The joke is not subtle, but it is well built: the hat establishes the witch, the broom establishes the vehicle, and the serif capitals deliver the punchline with a straight face. The crescent moon and stars are decorative, but they also do the work of setting: this is a night scene, a Halloween scene, a scene where a broom is a legitimate mode of transport.
What interests me is the way the design withholds the joke until the eye has traveled the whole composition. The hat sits at the top, the words below it, the broom to the right. A viewer who reads only the text might see a mundane boast about driving a manual car. A viewer who sees only the hat and broom might see a generic witch graphic. The shirt works because it forces both readings to occupy the same space. That is a more deliberate piece of design than the Halloween shirt market usually gets credit for. The engraved style of the hat — cream lines, a square buckle, a curved brim — gives the image a slightly antique, almost woodcut quality, which makes the pun feel drier than it would in a cartoonish font. The serif capitals match that dryness. Nothing here is winking. The shirt delivers the joke deadpan, and that is why it lands.
I have written before about the way food and drink packaging turns words into costumes — the label that says salted because the law requires it, the pun that sells a shirt because the pun is the product. This shirt belongs to a different shelf: the seasonal costume humor shelf, where the joke is not about what you eat or drink but about what you can operate. Still, the mechanism is the same. A phrase is lifted from one context — the garage, the driver’s seat, the manual transmission — and dropped into another — the witch’s broom, the Halloween party, the October shirt. The pleasure is in the collision. The shirt does not need to explain the collision. It just needs to arrange the hat, the words, and the broom so that the collision happens in the viewer’s head.
The retail title uses the word women, and the listing describes the shirt as a women’s Halloween shirt. I am not going to make claims about fit or fabric or how it wears, because I have not worn it and the listing does not give me a textile specification I can verify. What I can say is that the design itself is not gendered in any obvious visual way. The hat, the broom, the moon, the stars, the serif capitals — none of these are inherently feminine or masculine. The pun works for anyone who has ever driven a manual transmission or wished they could. The title’s framing is a marketing decision, not a visual one, and it is worth separating the two. The artwork is the artwork. The title is the title.
There is a small editorial matter to state plainly: this publication and Quirky Cotton Collective share an owner. This is not an independent review. I am writing about the shirt because the artwork is the occasion for the piece, not because I am recommending a purchase. The observation that anchors this column is the design itself — the cream hat, the curved brim, the square buckle, the serif capitals, the long broom on the right, the crescent moon and stars and flourishes. That is what I have to work with, and it is enough for the argument I want to make.
The argument is this: the shirt is a small lesson in how a pun becomes a picture. The words Yes I Can Drive a Stick are not funny on their own. They become funny when the hat and the broom reframe them. The design does not rely on a caption or a speech bubble or an explanatory line. It relies on composition — the vertical stack of hat and text, the broom running down the side like a margin note, the moon and stars filling the empty space with atmosphere. The shirt is a visual sentence, and the grammar of that sentence is the joke. You read the hat, then the words, then the broom, and the meaning shifts between each element. That is a more sophisticated piece of design than it first appears, and it is worth pausing on before the season passes.
The shirt is listed at the Quirky Cotton Collective Amazon listing, where the artwork and title can be seen together. I am not going to tell you to buy it. I am going to tell you to look at it — at the way the hat sits above the words like a heading, at the way the broom runs down the right side like a column rule, at the way the moon and stars turn a plain background into a night sky. The design is doing more work than the pun alone, and that is the part worth noticing.
Return, then, to the document: a cream hat, engraved-style, pointed, with a curved brim and a square buckle. Below it, YES I CAN DRIVE A STICK in serif capitals. A long broom down the right side. A crescent moon, small stars, flourishes. The shirt is a Halloween object, but it is also a small piece of typography and layout, and it rewards the same kind of attention I would give a menu or a price board or a recipe card. The joke is the hook, but the composition is the craft. That is what I see when I look at it, and that is what I wanted to put down here before the season turns.
The artwork is a closed loop. The words I’M JUST HERE FOR THE BOOS run in a curve around a ghost who wears sunglasses and holds two foaming beer steins. The lettering frames the figure; the figure fills the lettering. There is no background, no bar, no graveyard, no pumpkin. Just the pun, the ghost, and the beer. That economy is the design’s whole argument, and it is worth reading closely before the shirt gets folded into a drawer in early November.
The pun itself is a homophone swap: boos for booze. The ghost supplies the first meaning; the steins supply the second. Neither half works alone. A ghost without beer is a generic Halloween decoration. Beer without the ghost is a Friday afternoon. The shirt’s joke depends on the two being locked together inside the same curved sentence, so that the eye reads boos and the brain hears booze at the same moment it registers the foam. That is a small piece of graphic engineering, and it is the reason the design reads instantly at a distance — the way a good menu board reads from the door.
The ghost is the interesting part. It is not a scary ghost. The sunglasses do the work of defusing it: this is a ghost on its day off, a ghost that has decided the afterlife is a social occasion. The two steins are held out, one in each hand, in the posture of someone arriving at a party and looking for the person who invited them. The foam is drawn as a scalloped cap on each glass, which is the standard visual shorthand for cold beer — the same shorthand a tavern sign uses. None of this is subtle, and it is not meant to be. The design is a costume in the theatrical sense: it tells the room what role the wearer is playing before anyone speaks.
What the shirt does, structurally, is borrow the grammar of the bar and the grammar of the ghost story and run them through the same sentence. I’m just here for the boos is a line a person might actually say at a Halloween party — the disclaimer of someone who came for the drinks and not the haunted house. The shirt prints the disclaimer on the chest. That is a familiar move in food-and-drink apparel: the wearer outsources the small talk to the garment. The garment says I am here for the beer, the ghost is the excuse, and the wearer is freed from having to explain why they are standing by the cooler.
I am not going to claim the shirt is comfortable, or that it fits well, or that it survives a wash cycle. I have not worn it. What I can read is the design as supplied: the curved lettering, the sunglasses, the two foaming steins, the ghost. The retail title calls it I’m Just Here For The Boos Funny Ghost Beer Halloween, and the visual observation confirms the wording and the composition. The shirt is sold by Quirky Cotton Collective; you can find the listing here. That is the object. The rest is interpretation, and I will keep it labeled as such.
The interpretation is this: the design is a piece of seasonal signage that happens to be wearable. It works the way a good chalkboard works — one image, one line, no clutter. The ghost is the mascot; the beer is the product; the pun is the hook. In a category crowded with skeletons, pumpkins, and black cats, the ghost-with-beer is a specific choice, and the specificity is what makes it legible. A generic Halloween shirt says October. This one says October, and I would like a drink. That is a narrower audience, and a more honest one.
There is a small craft point worth naming. The lettering curves around the figure rather than sitting above or below it. That is a harder layout than a straight line, because the words have to stay readable as they bend. The designer has kept the phrase short enough to survive the curve — five words, no subordinate clauses — and has placed the ghost at the center so the eye lands on the face first and the sentence second. The sunglasses are the anchor: they give the ghost a direction of gaze, even though the eyes are hidden. The steins are held symmetrically, which keeps the composition balanced and lets the foam read as two matching exclamation points. None of this is accidental, and none of it is precious. It is the kind of layout a sign painter would recognize.
The shirt is not a document. It is not a menu, a ledger, or a payroll record, and I am not going to pretend it is. But it belongs to the same family of objects this column keeps returning to: the things people make to say who they are in a room. A menu tells you what a kitchen values. A shirt like this tells you what a person values on a Saturday night in October. The values here are simple — beer, company, a ghost who is not trying to frighten anyone — and the design states them without apology. That is worth a second look before the season turns and the shirt goes back in the drawer.
Editorial note: This publication and Quirky Cotton Collective share an owner.
In 1910, a Chicago publisher called the Hotel Monthly Press issued a working manual by Charles Fellows titled The Menu Maker: Suggestions for Selecting and Arranging Menus for Hotels and Restaurants. It is not a history. It is a tool for stewards, headwaiters, and chefs, and its index is the most useful thing in it for anyone trying to understand what the American Plan was. The index lists, in sequence: “Breakfast, American Plan,” “Luncheon, American Plan,” “Dinner, American Plan,” then “Luncheon, European Plan,” “Dinner, European Plan.” Two formats, side by side, in one book. No argument between them is recorded on those pages. What the manual records is a difference in how a bill of fare is built.
That distinction matters because the American Plan is often described as a pricing scheme — your room came with meals — and then treated as a historical curiosity or a marketing gimmick. The 1910 manual shows something more concrete. The plan was a menu format, and a menu format is a labor document. It tells you how many dishes a kitchen was expected to produce, how often the card changed, and who absorbed the cost of variety.
Fellows prints a full American Plan breakfast card for Sunday, March 1. The card is organized by category: fruit, cereals, eggs, hot meats, specials, potatoes, hot breads, coffee and tea. Under cereals alone the card lists Pillsbury’s Vitos, Shredded Wheat Biscuit, Cream of Wheat, Apitezo, oatmeal, boiled rice, Pettijohn’s Breakfast Food, Force, puffed rice, toasted corn flakes, Zest, Malta Vita, and Grape Nuts. Under eggs: boiled, fried, scrambled, poached or shirred, au gratin, à la Creole, au beurre noir, and omelettes plain, ham, parsley, jelly, or Spanish. Under hot meats: ham or bacon, calf’s liver and bacon, veal cutlet, pork, mutton, veal or lamb chops, tenderloin steak, sirloin steak, rump steak, with garnitures. Then a “special” column of starred dishes that the manual says are “changed from meal to meal.”
Read as a menu, that is a list of choices. Read as a labor document, it is a schedule. A kitchen producing that card needed a steward to order against it, a chef to assign it, and a brigade to execute it before the dining room opened. The manual’s own note on the starred dishes — changed from meal to meal — is the part that matters most. Variety on an American Plan card was not a one-time flourish. It was a recurring obligation, and the labor behind it recurred with it.
Fellows does not describe the wages, hours, or working conditions of the people who produced these cards. That silence is itself part of the record. The manual is written for the people who designed the work, not the people who did it.
In the same manual, the European Plan luncheon and dinner cards print a price beside each dish. Rockaways, half-dozen, 35 cents. Broiled live lobster, chili sauce, $1.10 or 60 cents. Mashed potatoes, 15 cents. Pot of coffee for one, 15 cents; for two, 25 cents. The American Plan cards in the same book carry no per-dish prices at all.
That is the visible difference between the two formats as Fellows presents them: one card prices the dish, the other does not. It is a format difference, and it is worth being precise about what it does and does not prove. It does not prove that hotels were at war over the two plans. It does not prove that American Plan guests ate better or worse than European Plan guests. It shows that a menu maker working in 1910 had to be able to build both kinds of card, and that the two kinds required different accounting at the point of service.
For a modern reader, the analogy is formal rather than historical: a fixed-price breakfast and an à la carte menu still differ in where the price appears and who carries the risk of the guest’s choices. The 1910 cards are a clean teaching example of that difference, nothing more.
The manual’s index also lists “A Dissertation on The Bill of Fare Maker vs. The Waiter and the Tipping System,” at page 57. That line is the only part of the dissertation retrieved for this piece. It is enough to establish that Fellows’ book treated the relationship between menu design, waiters, and tipping as a subject worth a chapter. It is not enough to say what the chapter argues. Anyone citing Fellows on tipping should read page 57 first.
What the index line does suggest, cautiously, is that the American Plan was understood by at least one trade publisher as a labor-relations question and not only a pricing question. A card without printed prices changes the waiter’s role: the guest is not choosing against a visible number, and the transaction between guest and server is structured differently. Whether Fellows thought that improved or worsened the waiter’s position is a question the index cannot answer.
Twenty-three years after Fellows’ manual, The New Yorker published a portrait of the Cheshire House in Keene, New Hampshire, in its issue of February 11, 1933. The piece describes a country hotel where the bill of fare carried no prices. The narrator, having forgotten to ask at the desk, reports: “As it turned out, you just eat—for two dollars.”
That is a surviving American Plan practice, described by a visitor. It is one hotel, in one year, in one region. It is not evidence of a national trend, and it should not be used as such. But it is useful for a different reason: it shows what the American Plan looked like at the point of service, and it shows who appears in the record when a writer describes that point of service.
The waitress in the piece is asked how long she has worked there and is “noncommittal.” She is asked how much dinner costs and says she doesn’t know. The narrator suspects she does know and is minding her own business. Either way, the person carrying the plate is not the person who holds the price. The piece also notes “the aged Negro bellboys who sit on a bench in the lobby,” and, on a tour of the kitchens, “a bent and aged party in an apron” cutting up masses of liver for cats. These are asides. They are also the only labor the document records in any detail.
The evidence assembled here supports a modest argument. The American Plan, as it appears in a 1910 trade manual, was a menu format with a recurring labor obligation: a daily-changing card, no printed prices, and a kitchen organized to produce variety on schedule. The European Plan, in the same manual, was a format with printed prices per dish. The two coexisted in one working book. A 1933 magazine portrait shows one American-plan hotel still operating on the no-prices convention, with its labor visible mainly in passing.
The evidence does not support a story about hotels fighting over the two plans. No retrieved source documents such a conflict. It does not support claims about the plan’s rise, decline, or market share. It does not support a summary of Fellows’ tipping chapter. And it does not support treating the Cheshire House as typical of American hotels nationally in 1933.
What it does support is a method. When a menu card carries no prices, the historian’s question should shift from what the guest paid to who produced the card and under what arrangement. The 1910 manual answers the first half of that question in detail and leaves the second half blank. The 1933 portrait fills in a little of the second half, in asides, about a waitress, some bellboys, and a person cutting liver. That imbalance — the menu fully documented, the labor barely — is the most honest thing the sources have to say about the American Plan.
Was the American Plan the same as full board? In the sources here, the American Plan appears as a menu format in which the bill of fare carries no per-dish prices. The 1933 Cheshire House portrait describes a flat two-dollar dinner with no prices on the card. The retrieved documents do not support a broader definition covering every meal arrangement at every hotel.
Did hotels stop using the American Plan because of competition with the European Plan? Nothing in the retrieved sources documents that. The 1910 manual treats both as formats a menu maker might need. The 1933 portrait shows one hotel still using the no-prices convention. A claim about why the plan declined would need dated ledgers, contracts, or trade-association proceedings that are not in this evidence set.
Where can I read the primary documents? Fellows’ Menu Maker (Hotel Monthly Press, 1910) is available in full text through the Internet Archive, catalog record cu31924000500433. The Cheshire House portrait appeared in The New Yorker on February 11, 1933.
What would settle the open questions? A dated hotel ledger or payroll record showing American Plan labor arrangements; a price board or contract from a specific hotel showing how the plan was priced in practice; and the full text of Fellows’ page 57 on the bill of fare maker, the waiter, and tipping. Until those are in hand, the argument stays where the documents put it: a menu format, a labor obligation, and a record that names the work mainly in asides.
The garment preview shows two lines of white type on a dark shirt. The first line reads IN DOG BEERS. The second reads I’VE ONLY HAD ONE. There is no illustration, no paw print, no bottle, no cartoon dog. The joke is carried entirely by the letterforms and the space between them. That restraint is the first thing worth noticing, because it tells you what kind of object this is: a typographic pun, not a costume.
The phrase works by substitution. “Dog years” is the familiar idiom — the notion that a dog ages faster than a human, so one calendar year counts for several. The shirt swaps years for beers and lets the arithmetic do the rest. If one dog beer equals several human beers, then the speaker who has “only had one” is making a claim that collapses under its own logic. The joke is not that the wearer drinks a lot. The joke is that the measurement system is rigged. It is a unit-conversion problem disguised as a drinking confession.
That distinction matters, because the easy reading of a shirt like this is the wrong one. A lazy interpretation would treat it as generic beer merchandise — a logo for people who like drinking, a conversation starter for a bar. But the design does not show a beer. It does not name a brand, a brewery, a style, or a quantity in pints. What it shows is a sentence about scale. The humor lives in the gap between the two lines, where the reader has to supply the missing term — dog years — and then watch the sentence fall apart. The shirt is closer to a riddle than to a slogan.
I read the type as plain and unadorned: white letters, all capitals, stacked in two centered lines. The visual observation does not specify a typeface, and I will not invent one. What I can say is that the design withholds decoration. There is no distressed texture described, no drop shadow, no banner, no frame. The dark ground does the work of contrast, and the white letters do the work of statement. This is a design that trusts the sentence to be funny on its own. Whether it succeeds is a matter of taste, but the choice is legible: the words are the object.
There is a small grammatical pleasure in the first line, too. “In dog beers” is not a phrase anyone says. It is a deliberate malformation, a hinge that signals the joke before the second line arrives. The reader who knows the idiom fills in the correction automatically — dog years — and then has to re-read. That momentary stumble is the mechanism of the pun. The shirt is not asking you to laugh at a dog or at a drinker. It is asking you to notice a category error and enjoy it.
For cooks and people who work in kitchens, the shirt has a particular resonance that has nothing to do with drinking. Kitchens run on measurement systems that are just as arbitrary as dog years. A “pinch” is not a unit. A “glug” is not a unit. A “medium heat” is not a temperature. The language of the line is full of substitutions that only make sense inside a shared frame — the same way “one dog beer” only makes sense if you accept the premise. The shirt is a small joke about the instability of units, and anyone who has ever tried to write a recipe for someone else knows how unstable units can be.
That is the argument I want to make about this object: it is funnier as a piece of language than as a piece of merchandise. The design does not lean on a picture of a dog or a glass. It leans on the reader’s familiarity with an idiom and the reader’s willingness to do the arithmetic. The shirt is a dog beers shirt in the literal sense that the words are printed on a shirt, but the phrase is doing the work, not the garment. The dark ground and white letters are a delivery system for a sentence that only lands if you read it twice.
I am not going to tell you it is comfortable, or that it fits well, or that people will stop you on the street. I have not worn it. What I can tell you is what the preview shows: two lines of white type, centered, on a dark shirt, with no illustration. That is the whole design. The rest is reading.
And reading it again, I notice that the second line is quieter than the first. “I’ve only had one” is a defensive statement, the kind of thing someone says when they suspect they have had more than one. The first line sets the trap; the second line springs it. The shirt is structured like a two-beat joke, and the pause between the lines is where the laugh is supposed to happen. Whether it does is up to the reader, but the structure is sound.
The object is dated by the retail observation of October 14, 2026, and the listing identifies the brand as Quicky Cotton Collective. That is the extent of what the record supports. The design is a typographic pun, and the pun is about the mismatch between two systems of counting. The shirt does not resolve the mismatch. It just puts it on a dark ground in white letters and lets you do the math.
Editorial note: This publication and Quicky Cotton Collective share an owner.