Yomimaru Logo
Library Pricing
AI Study Tools

Japanese Grammar AI Explainer: What Maru-Sensei Tells You and When to Use It

By Yomimaru Sensei Published on 2026-06-25
Japanese Grammar AI Explainer: What Maru-Sensei Tells You and When to Use It Illustration

Maru-Sensei is a Japanese grammar AI explainer built into Yomimaru's reading interface. Highlight any sentence in any text you are reading — from Yomimaru's library or a document you imported — and Maru-Sensei returns a real-time structural breakdown: what each particle does in that sentence, where embedded clauses begin and end, what the dropped subject is, and what the conjugation chain means. It works on any text, not just pre-annotated content.


The Gap That Grammar Books Cannot Fill

Every serious Japanese learner has at least one grammar reference: Genki, Tae Kim, A Dictionary of Japanese Grammar, or Bunpro. These are genuinely useful tools for learning what te-form is, what the passive construction looks like in isolation, and what conditionals mean in theory.

The gap opens when you sit down with a real Japanese sentence and cannot parse it — even though you know all the individual grammar patterns it contains. A sentence like:

「長年その問題を避けてきたとされる政府の対応が、ようやく見直されつつあるとの報告が出た。」

You might recognise every word. You might recognise passive, te-form, and quotation particles. And yet the sentence is opaque, because the grammar is not the problem — the structure is. Which clause modifies which noun? Who is the actual subject? Where does the main clause start?

No grammar dictionary answers this. They all describe patterns; none of them parse a specific sentence and tell you how those patterns fit together in this instance. That is what Maru-Sensei does.


What Maru-Sensei Actually Tells You

Particle disambiguation in context

Japanese particles carry the grammatical structure of a sentence — は marks topic, が marks subject, を marks object, に marks direction or receiver. But in complex sentences, it is not always obvious which noun is connected to which verb via which particle.

Maru-Sensei identifies what each particle is doing in the specific sentence you highlighted — not in the abstract, but in context. It tells you which が is the subject of which embedded verb, and which は is contrasting which items.

Embedded relative clause structure

Japanese modifies nouns by placing entire clauses in front of them, with no joining word. A single noun can be modified by three stacked clauses, each with its own verb and particles. Maru-Sensei identifies where each modifying clause begins, which noun it ultimately modifies, and what relationship it establishes.

This is the single most common reason intermediate learners cannot parse N2 and N1 sentences. The clauses are all recognisable in isolation; the stacking pattern is not.

Dropped subjects and implicit speakers

Japanese regularly drops the subject of a sentence when it can be inferred from context. In a long text, the implicit subject can shift mid-paragraph without any explicit pronoun. Maru-Sensei identifies what the dropped subject is based on the surrounding text context, so you know who is doing what.

Passive and causative constructions

Japanese passive (〜れる/〜られる), causative (〜せる/〜させる), and causative-passive (〜せられる/〜させられる) stack onto verbs and change who is acting on whom. When these appear in formal written contexts — newspapers, business reports, academic writing — they are compressed and easy to misread.

Maru-Sensei identifies which construction is active and explains who the agent and patient are in that specific sentence.

Conditional nuances

Japanese has four main conditional forms — 〜たら, 〜ば, 〜と, 〜なら — each with subtle differences in implication and usage. Standard grammar explanations describe the difference in theory. Maru-Sensei identifies which conditional is present and explains what specific implication it carries in that context: is this a neutral if-then? A hypothetical? An expectation? A recommended condition?

Formal written register patterns

N2 and N1 texts use written-language grammar patterns that almost never appear in spoken Japanese: 〜にすぎない, 〜にもかかわらず, 〜をめぐって, 〜に基づいて, 〜に過ぎない. These compound expressions function like single grammatical units.

Maru-Sensei identifies them as units, explains what they mean as a whole, and places them within the sentence structure so you understand how they connect to the main argument.


How Maru-Sensei Differs from Other Grammar Tools

| | Grammar dictionaries | Bunpro | DeepL / Google Translate | ChatGPT (standalone) | Maru-Sensei |

|---|---|---|---|---|---|

| Explains patterns in isolation | ✓ | ✓ | ✗ | ✓ | ✓ |

| Works on a specific sentence | ✗ | ✗ | Translation only | ✓ | ✓ |

| Integrated into reading interface | ✗ | ✗ | ✗ | ✗ | ✓ |

| Works on custom imported text | N/A | N/A | ✓ | ✓ | ✓ |

| Preserves reading flow | ✗ (switch apps) | ✗ (switch apps) | ✗ (switch apps) | ✗ (switch apps) | ✓ (inline) |

| Context-aware (reads surrounding text) | ✗ | ✗ | ✗ | Sometimes | ✓ |

The critical difference is the last two rows. Every other tool requires you to leave what you are reading, copy the sentence somewhere else, and then return — breaking the reading flow every time. Maru-Sensei is integrated directly into the Yomimaru reading interface, so the explanation appears alongside the text without any context switch.


How to Use Maru-Sensei Effectively

When to use it:

  • You have read a sentence twice and still cannot determine who is doing what to whom.
  • You recognise all the vocabulary but cannot figure out how the clauses connect.
  • You see a formal written construction you have not encountered before.
  • You want to confirm whether your grammar reading of a sentence is correct before moving on.

When not to rely on it:

  • As a substitute for reading. Using Maru-Sensei on every single sentence prevents fluency from developing. Use it for sentences that genuinely block comprehension.
  • For vocabulary lookups. Tap any individual word for the built-in dictionary — Maru-Sensei is for sentence-level structure, not word meanings.

The most effective workflow:

  1. Read a sentence. If you can extract the meaning, continue.
  2. If the sentence blocks you, attempt to parse it yourself first — identify the main verb at the end, work out the basic subject-object-verb frame.
  3. Then highlight and trigger Maru-Sensei. Compare its breakdown to your attempt.
  4. Re-read the sentence with the correct structure in mind. Your brain internalises the pattern.

This active-then-verify loop builds grammar intuition faster than looking up every sentence passively.


By JLPT Level: What Maru-Sensei Helps With Most

N4–N3 learners: Embedded relative clauses, dropped subjects, te-form chains connecting multiple actions, basic passive recognition.

N2 learners: Stacked modifying clauses in formal texts, perspective markers and hedging language, compound conditional structures, formal written vocabulary that functions as grammatical units.

N1 learners: Abstract essay structures where the author's argument is layered across multiple paragraphs, formal honorific and humble form chains in business contexts, academic hedging with nominalisation, sentence-final nuance particles in literary prose.


Maru-Sensei as a Grammar Pattern Database

Every explanation Maru-Sensei generates is tied to the sentence you highlighted. Over a reading session, you may encounter the same construction three or four times across different sentences. Each time Maru-Sensei explains it, you are building a mental corpus of that pattern in real usage — not in isolation.

This is fundamentally different from studying grammar lists. Grammar lists tell you what a pattern is; reading shows you what it does. Maru-Sensei closes the gap between the two.

Try Maru-Sensei on any Japanese sentence →


Further Reading

Frequently Asked Questions

What is Maru-Sensei in Yomimaru?

Maru-Sensei is the AI grammar explainer built into Yomimaru's reading interface. When you highlight any sentence in a text you are reading — whether it is from Yomimaru's library or a document you imported yourself — Maru-Sensei returns a real-time structural analysis: which particles do what, where embedded clauses begin and end, what the dropped subject is, and what the verb conjugation chain means.

How is Maru-Sensei different from a Japanese grammar dictionary?

Grammar dictionaries like A Dictionary of Japanese Grammar explain grammar patterns in isolation — te-form, passive, conditionals. They require you to already know which pattern you are looking at before you can look it up. Maru-Sensei works on the actual sentence in front of you, identifies the specific patterns active in it, and explains them in the context of that specific sentence. You do not need to identify the pattern first.

Can Maru-Sensei explain grammar in texts I import myself?

Yes. Maru-Sensei works on any text you paste into Yomimaru — a news article, a novel chapter, a manga transcript, a business email. It is not limited to pre-annotated content in a curated library. This is the key difference from platforms like Satori Reader, which provide static pre-written grammar notes only on their own content.

What grammar topics can Maru-Sensei explain?

Maru-Sensei can explain any Japanese grammar structure it encounters: particle disambiguation (は vs が in a specific sentence), embedded relative clauses (who is modifying what noun), passive and causative constructions (who is acting on whom), dropped subjects (who is performing the action based on context), conditional forms and their nuances, and formal written register patterns common in N2 and N1 texts. It provides explanations in plain English, not grammar jargon.

Ready to Read Authentic Japanese?

Ditch the simplified textbooks. Yomimaru makes real Japanese light novels, blogs, and news fully readable for N5 to N1 learners with instant dictionary lookups and AI-powered grammar support.