Describe a Picture in Japanese: A Practice Drill With Five Worked Examples

The frame in action: one picture, four sentences

Here is the whole drill on one photo of a station platform. The first sentence names the place with desu. The second says who is there with imasu. The third says what they are doing with te imasu. The fourth gives the feeling with sō desu. You could stop there in an exam and have done the task; everything else in this post is about making those four sentences easy and then adding to them.

The four questions

The frame is a set of four questions you ask yourself, in order, every time. Each question has one grammar pattern as its answer, so once you know the four patterns you never have to decide how to start. The questions are also exactly what a teacher or examiner asks when you dry up, so learning them means you understand the prompt too.

Grammar the drill needs

Four patterns carry the frame, and two more sharpen it. Imasu and arimasu say what exists and take ga. The te imasu form says what is happening now; it is also used for states like clothes (bōshi o kabutte imasu, wearing a hat). Sō desu attached to an adjective stem means looks or seems, from what you can see. To omoimasu adds your opinion. Position words tell the listener where to look; put them before the thing with ni.

Note the particle shift in the third sentence. The man was introduced with ga; once he is the topic you describe, he takes wa. New information takes ga, known information takes wa. In a four-sentence description that usually means ga in sentence two and wa in sentence three.

Miemasu, can be seen, is the natural verb for things in the distance. Oku ni yama ga arimasu is correct; oku ni yama ga miemasu sounds like someone looking at a picture.

Five practice pictures, described

Here are five ordinary pictures described in the frame, four sentences each. Read each set aloud, then cover the Japanese and rebuild it from the English. The pictures are deliberately dull: a café, a park in autumn, a kitchen, a classroom and a platform at night. The drill is for ordinary scenes, because exams and daily life are full of them.

Two forms deserve a note. Ii becomes yosasō (looks good), so kimochi ii becomes kimochi yosasō. Tsukarete iru, to be tired, is a te-iru state, so looks tired is tsukarete isō desu, not tsukaresō desu, which would mean looks like they will get tired.

A park in autumn, labelled

The scene below shows the second picture with its labels, so you can see how the four sentences map onto what is actually in the frame. Describe it yourself before reading the labels, then compare. If you said something the labels do not, that is fine; pictures have more in them than four sentences, and the frame is a floor, not a ceiling.

A daily drill with your own photos

The drill becomes a habit when the pictures are yours. Your phone already holds hundreds of scenes you know well, and describing a photo you took is easier and more interesting than describing a textbook picture. One photo a day, sixty seconds, out loud, in the frame order. The checklist keeps it honest.

• Pick a photo without choosing: the most recent, or scroll and stop at random.
• Start the timer, then say the four sentences in order: place, people or things, action, feeling.
• If a word is missing, say nanika (something) and keep going; look it up afterwards.
• Once a week, write one photo's four sentences down and get them corrected.
• Every second week, add a fifth sentence that guesses something: the season, the time, what happened next.

The nanika rule matters. The drill is about producing four sentences without stopping, and a missing noun is not a reason to stop. Learners who pause to look up every word never get to sentence four.

Levelling up: guesses and feelings

When four sentences come easily, add the two kinds of sentence that make a description sound like a person rather than a caption. Guesses use deshō (probably), kamoshiremasen (might be) and to omoimasu (I think). Feelings use kanji ga shimasu, it has a feeling of, and your own reaction. These are also what examiners reward and what a friend showing you a photo actually wants to hear.

Mistakes in picture descriptions

• Arimasu for people. A woman is onna no hito ga imasu; the same woman's bag is kaban ga arimasu.
• Wa for new things. The first mention of each person or object takes ga; wa comes once they are the topic.
• Isogashii desu for looks busy. If you are judging from the picture, say isogashisō desu.
• E instead of shashin. E is a painting or drawing; a photo is shashin, and most exam and app pictures are shashin.
• Stopping at a missing word. Say nanika and keep the four sentences moving.

A partner for the drill

The drill works alone, but it works better with someone who asks the questions. Mocchi, Unihongo's study assistant, can set the picture for you: ask it to describe a simple everyday scene in English, describe it back in Japanese in four sentences, and let Mocchi correct the sentences and ask one follow-up. Because it keeps track of what you keep missing, it will notice if ga and wa keep swapping. Later, /correct takes a weekly written description.

For the spoken version, hold a photo in front of you in AI Sensei, tell Haru-sensei what is in it, and ask Haru-sensei to keep asking what else Haru-sensei should know about it. Both start with a free account at unihongo.com/login, and the first picture can be whatever is on your phone right now.