A Japanese graduate research lab — kenkyushitsu — is not a workplace you walk into at ten and leave at six. It is a small vertically organized community with its own weekly rituals, its own seasonal events, its own etiquette around hierarchy and reciprocity, and its own unwritten expectations about presence. Foreign students who arrive expecting a Western-style research group and adjust late tend to spend their first year confused and their second year exhausted. The applicants who thrive are the ones who understand the cultural shape of the lab before they accept the offer. This 2027-stamped guide walks through what daily life inside a Japanese lab actually looks like — the zemi, the senpai dynamics, the work hours, the drinking culture, the summer camp — and how international students can navigate it without losing the research years.
The lab as a vertical community, not a job
In a Western university, "the lab" usually means a research group: a professor, some postdocs, and a rotating cast of graduate students who collaborate but go home to separate lives. The professor lives in their faculty office down the hall; students have desks in shared graduate rooms; people drift in and out on their own schedules. In Japan the lab is a single physical and social unit. Every member — undergraduates writing their senior thesis, master's students, doctoral students, the professor, the assistant professor, sometimes a technician or two — shares one large room or one connected suite of rooms. You get a desk in your first week and you keep that desk until you graduate. You eat lunch nearby with the lab. You attend the weekly seminar that the entire lab sits through. You are part of a community whose membership is defined as much by your physical presence as by your academic output.
The professor is not a remote figure who appears at meetings; they sit in an attached private office, often with the door open, and the social weather of the lab orients around their presence. The senpai (senior students) are not abstract collaborators; they are sitting two desks over, watching whether you turn up at 10 am or 1 pm, whether you stay for the late-afternoon zemi practice, whether you join the dinner after a paper is submitted. This communal structure is the core of the Japanese lab system, and almost everything else in this guide flows from it. If you choose to read only one sentence of this page before you accept an offer, it should be this: in Japan, the lab is a small village you live in for two to five years, not an office you use for research.
This shape changes how you choose a lab in the first place. The seven-dimension scorecard in how to choose a Japanese graduate lab treats culture and senpai dynamics as first-class evaluation criteria, not soft factors. By the time you have read both that guide and this one, you should be able to look at any lab page and predict, within reason, what your daily life there will feel like.
The zemi: the weekly seminar that organises everything
The zemi (ゼミ) — sometimes written セミナー or 研究会 — is the most important recurring event in a Japanese lab. Almost every active research lab holds one weekly. It is two to four hours long, attendance is mandatory, and one or two students present their research progress to the whole group while the professor and senpai ask questions. In smaller labs everyone presents on rotation; in larger labs the schedule rotates every few weeks. There is no chance to skip in a healthy lab — if you miss zemi without illness, you owe the professor an explanation, and if you miss twice you have created a small social problem.
The zemi serves three purposes simultaneously. It is the technical review meeting where the lab works through your data, your plots, your method, your literature gaps. It is the social ritual where the lab sees who is producing and who is drifting. And it is the training mechanism through which junior students absorb what "good research" looks like in this lab — not from a syllabus but from watching senior students get praised or pushed back on. For an international student in 2027, the zemi is also the moment each week when the professor forms an updated mental image of your trajectory. Whether your slides are clean, whether you can answer questions in English when the rest of the lab uses Japanese, whether you cite the paper your senpai gave you last week — all of this lands in front of the professor in real time, and the impression compounds across months.
The format varies. Some labs run a paper-reading zemi in addition to the progress-report zemi: every member presents one recent paper from a top venue, summarises the contribution, and leads a discussion. Others alternate weeks between paper reading and project updates. Engineering and computer-science labs tend to lean progress-heavy; humanities and theoretical labs tend to lean paper-heavy. Either way, the rhythm is the same: a weekly forum where your work is visible to the whole community, and where the senpai network forms its opinions about you. The single most reliable way to thrive in a Japanese lab is to take the zemi seriously from week one.
Kohai and senpai: reciprocity is the operating system
Japanese labs run on the senpai (senior, 先輩) and kohai (junior, 後輩) relationship, and that relationship is the operating system of the lab. Senpai are anyone who entered the lab before you — even six months earlier counts, in a strict reading. In practice the social weight scales with the year gap: a doctoral student who started three years before you is a major senpai; a master's student in their second year when you arrive is a working senpai; an undergraduate who joined six months earlier is technically a senpai but operationally a peer. The professor is above this entire system; senpai/kohai dynamics describe student-to-student relationships specifically.
What senpai do for you, in your first year:
- Teach you the lab equipment, the analysis pipeline, the cluster login, the version of the code that actually runs versus the one in the README.
- Read your zemi slides the night before and mark up the embarrassing ones.
- Translate department announcements, scholarship deadlines, and university bureaucracy that arrives only in Japanese.
- Edit your manuscript drafts before they go to the professor — sometimes line by line.
- Explain the unwritten lab rules: who cleans the kitchen, when you can leave for the day, which seat is the professor's seat, whether it is acceptable to wear shorts in August.
What you do for senpai, also in your first year:
- Pour drinks for them at lab parties (this is real; the etiquette of pouring beer for seniors is one of the first social rituals you will absorb).
- Help set up the lab summer camp logistics, conference posters, equipment moves.
- Run small errands for joint experiments — fetch reagents, schedule equipment time, compile data.
- Take careful notes during zemi when seniors present, especially when the professor gives feedback you can later relay.
- Accept the role of "the new person" graciously: do not skip social events in your first six months unless you have a real reason.
The system is unpaid, unscheduled, and almost entirely informal. It is also the single biggest determinant of whether your day-to-day life in the lab feels supported or isolated. A lab where senpai are visibly mentoring kohai is a lab that will absorb you. A lab where senior students avoid juniors, eat alone, and never appear at group events is a lab where you will be on your own for two to five years. During a lab visit before accepting an offer, watch how senior and junior students interact when the professor is not in the room — that interaction is what your year two will actually look like.
Two years later, when you become a senpai, you do all of this in reverse. The new kohai pour your drinks, you read their slides, and the cycle continues. The reciprocity is the point. International students who try to opt out of one side of the system — accepting senpai help in year one, then refusing to mentor kohai in year three because "I am too busy" — quickly find themselves outside the lab's social core. For a more candid account of how this affects students who arrive without Japanese, see accepted into a Japanese lab without Japanese .
The PI: a different kind of supervisor
The professor — the principal investigator, in Western terminology — sits at the top of the lab hierarchy with significantly more authority than a typical Western PI but also, paradoxically, often less day-to-day involvement. A Japanese full professor controls admissions, funding allocation, authorship order, conference travel decisions, the speed of your degree, and the ultimate sign-off on your thesis. They cannot be appealed past in any meaningful sense; the department is largely ceremonial in academic decisions about you. In that sense the professor is more powerful than a Western PI.
In daily research mentorship, however, many Japanese professors are less hands-on than US or European counterparts. The dominant model is one-on-one meetings every one to three weeks (sometimes longer), plus their seat at the weekly zemi. They are less likely to read your full paper drafts in detail, less likely to do live whiteboard sessions on your method, less likely to message you on a weekend. The assumption is that senpai handle the close coaching and that the professor sets direction, judges output at zemi, and intervenes when needed. In a healthy lab this is fine; the senpai network compensates. In an unhealthy lab — small group, no strong senior students, distant professor — international students can go weeks without substantive supervision and only realise it months later.
There is a second category, especially in younger or international-facing labs: hands-on professors who meet weekly, co-author every student paper, and read full drafts. These are the professors most international students should target, particularly without prior independent research experience. The choosing a Japanese graduate lab guide covers how to identify hands-on supervisors before you commit. The professors directory cross-references published faculty profiles by field and supervision style where the data is available.
Work hours, presence, and the unwritten schedule
Quantitatively, a typical STEM lab in 2027 expects 50 to 60 hours per week, including evenings on lab premises. The shape of those hours matters as much as the count. Most labs do not enforce a 9 am start the way a corporate office would; the typical day looks more like arrive between 10 am and 12 pm, work through the afternoon, take a long communal dinner break around 7 pm, return for a few more hours, and leave between 9 and 11 pm. Weekends are nominally free but informally expected near conference deadlines, before zemi presentations, and during major experiment runs. Some labs are stricter: an explicit "core time" of 9 am to 9 pm Monday to Friday is not unusual at certain national universities, with the implicit assumption that weekends will see you in the lab when work demands it.
Lighter labs do exist. A growing minority of younger professors, especially those who themselves did postdocs abroad, run labs at closer to 40 hours per week with genuinely respected weekends. These labs tend to be in international-facing departments, often computer science or English-medium graduate programs. If protecting weekends and evenings is non-negotiable for you, you will need to filter hard during the application stage — see English-taught Master's in Japan 2027 and Computer Science Master's in Japan for the programs most likely to host these labs, and studying AI / ML in Japan for the field where lighter-hours labs are most concentrated.
Two cultural points are non-obvious to Western applicants. First, physical presence counts. A student who works from home four days a week and "logs the same hours" is not perceived as equally committed; the lab is a shared room, and being seen in it matters. The pandemic pushed Japanese labs toward more flexibility, but in 2027 the pre-pandemic norm has largely returned. Second, leaving before the senpai do is socially marked early in your tenure. The custom of nodding to the room before you pack up — in some labs, of saying osaki ni shitsurei shimasu ("excuse me for leaving ahead of you") — is a small reciprocal acknowledgment that you are part of a group. First-year international students who slip out without a word every evening are remembered as aloof. Three months in, this small ritual has become invisible second nature.
Lab membership: kitchen, room, events
Membership is concrete. You will share a lab kitchen — usually one large fridge, a microwave, an electric kettle, sometimes a rice cooker or coffee machine paid for from a previous year's lab budget. There is a cleaning rota; in many labs juniors handle Friday afternoon cleaning duty for the first year, with the rota rotating as new students arrive. There is a shared printer, a shared whiteboard, often a shared experimental area where time slots are negotiated person-to-person. Equipment etiquette is unwritten but strict: leave the bench cleaner than you found it, log the reagents you used, do not run the GPU cluster on a Friday night without telling the senpai whose paper deadline is Monday.
Lab events extend from the weekly through the seasonal. The weekly cadence is dominated by zemi. The monthly cadence often includes one nomikai (more on this below) plus occasional all-hands events — a sento (public bath) trip, a sports afternoon, a planning lunch. The seasonal cadence is structured: the kangeikai (welcome party) for new students in April, a hanami (cherry-blossom picnic) shortly after, the gasshuku in August, a smaller intermediate gathering in autumn, the bonenkai (year-end party) in December, and a soubetsukai (farewell party) for graduating students in March. International students who attend roughly two-thirds of these are read as part of the lab; those who attend almost none are read as visitors who happen to share a desk.
Drinking culture: nomikai and the ritual of the round table
Nomikai (飲み会) — drinking parties — are a recurring feature of Japanese lab life and remain so in 2027 despite a softer post-pandemic norm. Most active labs hold nomikai several times a year at minimum: a welcome party for new members, a year-end bonenkai, often a monthly izakaya night, and frequently extra rounds after paper acceptances or thesis submissions. The format is consistent: everyone meets at an izakaya around 7 pm, a single bill is paid at the end and split equally (warikan), the first round of drinks arrives together, the senior member raises their glass for kanpai, and the social conversation begins.
A few practical notes for international students:
- You do not need to drink alcohol. Oolong tea is universally available and unremarkable. Many Japanese students, especially in 2027, also order soft drinks without comment.
- Pouring etiquette matters in the first hour. When a senpai's glass is half-empty, offer to pour. They will pour for you. This is not optional politeness; it is the small reciprocal gesture that signals you are part of the group.
- Sitting position is loosely structured. Senior members and the professor (if present) sit at one end of the long table. Juniors sit nearer the entrance side or at the corners. You will be quietly directed in your first event.
- Stay through the first round at minimum. Leaving in the first thirty minutes is treated as a snub. Leaving after the first round (about ninety minutes in) is normal and uncontroversial; some students join a nijikai (second round) at a karaoke or bar, others go home. Either choice is acceptable once the first round is complete.
- Bring some cash, usually 3,000 to 5,000 yen, even if the splitting will happen by bank transfer afterward. The first-round bill is settled at the venue.
For students with religious or medical reasons not to drink, this is fully accepted in 2027 and worth declaring once early — the lab will adjust without comment. What is not accepted is skipping every nomikai for two years. The drinking events are where the lab decides whether you are part of the group, and the price of declining to attend is real even if no one tells you so directly.
The gasshuku: the lab summer camp
The gasshuku (合宿) is the annual lab retreat, usually two or three days in summer (most often August or early September), held at an onsen inn in the mountains, a seaside guesthouse, or a university-owned seminar facility. Costs are typically partly subsidised by the lab budget and partly paid by students — figure 10,000 to 25,000 yen per person, all in. The schedule is structured but informal: morning and afternoon research presentations, often longer-form than the weekly zemi (each student gets thirty to forty-five minutes to talk through their full project plan for the year), interspersed with onsen baths, group meals, walks, occasional sports, sometimes a shared late-night discussion of the lab's research direction over more drinks.
For incoming international students, the gasshuku is a high-leverage event. It is where senpai/kohai bonds formed during the spring become durable; where the professor sees you outside the desk environment for the first time; and where major lab decisions for the autumn — who works on which project, who is being prepared for which conference, who needs to step up in zemi — often get casually previewed between sessions. International students who attend their first gasshuku and then treat it as an annual fixture tend to be inside the lab's working core by year two. Those who skip in their first summer because of travel, internship abroad, or discomfort with the social format often spend year two trying to catch up on relationships that solidified without them.
Project assignment: less autonomy than American labs
A major cultural difference, often discovered late by Western applicants, is how research projects are assigned. In an American PhD, the dominant model is "find your own thesis" — students arrive, take coursework for two years, identify a research question, pitch it to their advisor, and work toward an independent dissertation. In a Japanese lab, projects are usually assigned. The professor has a portfolio of ongoing projects (often tied to specific KAKENHI grants or industry collaborations), and incoming students slot into one of those projects. You may have some choice among current projects; you rarely have the option to invent a new direction in your first year.
For master's students this is largely fine — two years is short, and a defined project gets you to a defended thesis on time. For doctoral students it is more consequential: your three-to-five-year research direction will be heavily shaped by whatever the professor's current grants are funding when you arrive. The healthy version of this looks like growing into the project, taking ownership of a sub-area by your second year, and steering the work toward your own questions by year three. The unhealthy version looks like remaining a technician on someone else's project for the entire degree. The difference is usually visible from the publication record: in healthy labs, the doctoral student is first author on substantive papers by the end of year two. See PhD in Japan: funding, duration, and English-taught options for how this interacts with funding cycles, and kenkyusei vs direct master's application for the choice many international applicants face about how to enter the system in the first place.
Output expectations: theses, papers, language
Output is structured around three documents in 2027. The master's thesis (shuushi ronbun) is universally required, defended in February of the second year. The doctoral thesis (hakushi ronbun) is required for PhDs, usually with a public defense in the final year and a committee that includes external examiners. Peer-reviewed papers are increasingly required, especially for doctoral degrees: most national universities now expect two to three first-author papers in journal-grade venues before allowing a PhD defense, and many top labs target more.
The language question — Japanese versus English — is shifting. In 2027, English is the dominant language of new research output in most STEM labs that produce internationally visible work. A lab that publishes its students' papers in IEEE, ACM, ACS, RSC, or Nature-family venues is publishing in English; that is also where international students will be expected to publish. Domestic Japanese journals and domestic conferences are still active and produce real research, but their share of output has been falling for a decade. For an international student in 2027 entering a research-active lab, you should expect to write your papers in English and potentially write your master's or doctoral thesis in English as well, depending on the program (some allow either; some require Japanese for the formal thesis even if the underlying papers are English). Confirm the thesis-language policy before you apply, particularly for English-taught programs covered in the English-taught Master's in Japan 2027 guide.
Foreign students: where the friction concentrates
The international students who run into trouble in Japanese labs are usually excellent on paper. The friction is rarely academic; it is cultural. In aggregate the recurring difficulties are:
- Work-hour expectations. Western students arrive expecting a 40-hour week with respected weekends, find the actual norm is 50–60 hours plus evenings, and either burn out trying to match it or get quietly marked as low-effort by keeping their original schedule.
- Group conformity. The instinct to schedule your own time, push back on assigned projects, or skip optional events reads in Japanese lab context as anti-team. The same behaviours that mark a "promising independent" American student mark a "difficult" foreign student in Japan.
- Japanese-language banter. Even in an English-medium lab, the side conversations, lunch chatter, zemi jokes, and gasshuku evening talks happen in Japanese. JLPT N3 grammar is enough for survival; following the actual social layer requires more, and the gap is most painful in the first six months.
- Drinking events. The optional-but-not-really nature of nomikai catches Western students used to "optional means optional." Skipping every event for two years quietly excludes you from the social network that determines a large fraction of your daily lab experience.
- Project autonomy. Students arriving from American or European systems sometimes resist the assigned-project culture for months, treating it as a problem to solve rather than the local norm. The resistance costs them mentorship time and slows their integration into the lab's actual work.
- Visible presence. Working from home, even productively, is perceived very differently in a Japanese lab than in an American one. Foreign students who maintain pre-pandemic remote habits often discover in year two that the senpai network has formed without them.
None of these are insurmountable. Most are manageable with awareness and modest adjustment. The students who fail to adjust are not the ones who hit one of these issues; they are the ones who hit four of them at once and did not know to expect any of them. For country-specific perspectives on this transition, see studying in Japan from the USA and studying in Japan from India .
How to navigate: practical tips for the first year
The mental model that travels best is: treat the first six months as ethnography. You are not yet trying to maximise research output; you are trying to learn how this specific lab actually works. Specific, durable advice:
- Show up. Be physically in the lab on a predictable schedule from week one. Late mornings are fine; absent days are not. The professor and senpai form their picture of you in the first three months and it is hard to revise later.
- Find your senpai. Identify one or two senior students you can ask small questions of repeatedly — equipment, etiquette, slide format. Their patience in your first month is the foundation of your second year.
- Attend the events. First nomikai, first gasshuku, first kangeikai. You can ramp down attendance later; you cannot retroactively be present. The arrival period is when membership is decided.
- Learn lab Japanese. You do not need fluency. You do need the twenty or thirty phrases that come up daily — greetings, food, zemi vocabulary, the small-talk that happens at the kitchen sink. Most international students massively underestimate how much social mileage two hundred hours of effort here produces. JLPT N3 is the practical floor for lab life; see the JLPT N3 study hub for a structured route.
- Take zemi seriously. Prepare slides early. Ask a senpai to glance at them the day before. Do not show up with overnight slides in your first semester; the impression is hard to undo.
- Pour drinks at parties. A small reciprocal gesture in the first nomikai is read as cultural awareness and earns disproportionate credit.
- Say goodbye when you leave. A nod and a quiet osaki ni in your first month is invisible by month three and signals that you understand the room.
- Ask before working remotely. If you want a half-remote schedule, negotiate it explicitly with the professor and the senpai network rather than just defaulting to it. Granted permission is fine; assumed permission is the friction case.
- Keep a part-time job within the rules. Most students supplement income with a tutoring or TA role; see working part-time as an international student in Japan for the legal limits and tax notes.
- Do the reciprocity in year three. When you become a senpai, mentor the new kohai actively. The system rewards visible reciprocation, and graduates whose juniors remember them well leave Japan with a network that compounds.
This bundle of habits is not optional polish; in a Japanese lab it is the substrate of the academic relationship itself. International students who treat these as the "real" job — not as cultural overhead distracting from research — discover a year in that the research has flowed naturally because the social architecture supporting it is intact.
Choosing well, before you arrive
Most of the friction discussed above is much easier to navigate if you chose a lab whose culture matches your tolerance levels in the first place. Some labs run softer hours, lighter event calendars, and more flexible remote norms; others run strict core time, mandatory attendance, and intense social schedules. Both are legitimate research environments; neither is right or wrong. The mistake is ending up in the strict version when you wanted the soft version, or vice versa, because you did not ask the right questions during the application stage. The how to choose a Japanese graduate lab guide and how to email a Japanese professor should be read together with this one. The first email and the lab-visit questions are where you actually learn what the daily culture will be — the lab page is almost always silent on the cultural shape, by design.
Funding is an adjacent constraint. The MEXT Scholarship 2027 complete guide covers the dominant route for international applicants. For specific timing on whether you should enter as a kenkyusei (research student) before formal admission or apply directly to the master's program, the kenkyusei vs direct master's application guide walks through the trade-offs — and worth noting, the kenkyusei period is also when you most directly observe the lab's culture before you fully commit, since you already sit in the lab room and attend the zemi.
Bottom line
The Japanese lab system is a vertically organised social community whose rituals — weekly zemi, senpai/kohai reciprocity, monthly nomikai, annual gasshuku, shared room and shared kitchen — are the substrate of the research itself, not separate from it. Foreign students who treat the cultural layer as overhead they would rather skip tend to stall in year two. Foreign students who treat the cultural layer as the first six months of fieldwork — pour the drinks, attend the camp, find the senpai, learn the lab Japanese, take zemi seriously, show up — tend to be productively embedded by the end of their first year and writing real papers by the end of their second. The research outcome and the social adaptation are not separable in the Japanese system; they are the same project. The applicants who internalise that before they accept an offer in 2027 are the ones who finish the degree on time, with strong recommendation letters, a senpai network that lasts decades, and a plausible path into either Japanese academia, the global research conversation, or the international tech and engineering hiring pipeline that increasingly recruits from Japan-trained PhDs.