Katakana Stroke Order: How to Write All 46 Kana and Why Angular Shapes Make It Easier

Katakana stroke order is the conventional sequence and direction for drawing each katakana. This teaching tradition runs through the Ministry of Education's 1958 guide 『筆順指導の手びき』. The MEXT (Japan's Ministry of Education) elementary curriculum requires first-graders to write hiragana and katakana in the prescribed order.¹²³

Katakana is mechanically lighter to handwrite than hiragana because its glyphs are lifted components of kanji, not full cursive simplifications. The same handful of universal principles therefore produce short, straight, mostly detached strokes.⁴⁵

Overview

The 46 base katakana each have a canonical stroke order rooted in the same MEXT pedagogy that governs kanji and hiragana.¹²⁶ Because each katakana is one component lifted from a man'yōgana kanji, a Chinese character used phonetically to write Japanese, the shapes are short, straight, and often detached. That makes the universal principles easier to see than in hiragana's continuous curves.⁴⁵

Once the principles are visible, the 46 per-kana diagrams stop being 46 things to memorize and become 46 instances of the same small ruleset.

What "stroke order" means for a katakana

A stroke (画, kaku) is one continuous mark made without lifting the pen. Lifting the pen ends the stroke and starts the next.⁷ Stroke order (筆順 hitsujun, also called 書き順 kakijun) is the conventional sequence and direction in which the strokes of a character are written.⁶

Direction is part of the order, not a separate property. A horizontal stroke is written left-to-right, and a vertical stroke is written top-to-bottom. A stroke drawn backwards is incorrect even when it lands in the correct slot of the sequence.⁶⁷

Most katakana are one-, two-, or three-stroke characters. The four-stroke outliers worth flagging early are ネ and ホ, both of which preserve the densely strutted skeleton of their source kanji.⁴⁸

Is there an official katakana stroke order?

Yes, by lineage. The de facto Japanese-education standard is the Ministry of Education's 『筆順指導の手びき』 (Hitsujun Shidō no Tebiki, "Guide to Stroke-Order Instruction"), published by 文部省 (Monbushō, MEXT's predecessor) in 1958.¹²⁶

The guide's primary scope is the 881 kyōiku kanji, the characters taught in elementary school, in regular script (楷書体). Its stated purpose is to unify the stroke order taught for the educational kanji.²⁶ In the same document, the guide explicitly extends its principles to kana and Roman letters: "このことは、漢字ばかりではなく、かな、ローマ字等についても、同じことが言える."² That lineage is what supplies the standard katakana stroke orders taught in Japanese elementary schools.²³

One canonical order, not the only legal one

The guide is explicit that other historically used orders are not declared wrong: "これを誤りとするものでもなく、また否定しようとするものでもない."² There is one canonical taught order per katakana, with documented tolerance for alternatives.

The MEXT elementary curriculum mandates in 第1学年 (first grade) that pupils "平仮名及び片仮名を読み、書くこと" and "筆順に従って文字を正しく書くこと": read and write hiragana and katakana following the correct stroke order.³ Katakana sits beside hiragana in the same first-grade requirement; the same 筆順 discipline applies to both.

How this article is organized

The article starts with principles, then gives a per-kana reference for all 46 grouped by gojūon row, the standard Japanese sound-order grid. It then explains why katakana is generally lighter stroke-wise than hiragana. The edge cases worth knowing are in the closing section. A reader who has already worked through the hiragana stroke-order article will recognize the scaffold; the rule contents shift to the katakana-specific applications.

The universal stroke-order principles applied to katakana

The MEXT guide organizes its rules into 2 大原則 ("great principles") and 8 原則 ("principles"). The same set governs both kanji and kana.²⁶ Katakana shapes draw most heavily on the first four (top-to-bottom, left-to-right, horizontal-before-vertical, enclosure-last), plus the dot/short-stroke convention and the three stroke-end types.

When two principles point in different directions, the hierarchy below is how the rule precedence falls out for katakana.

Top to bottom

大原則1 (上から下へ): write a character from its upper elements to its lower elements.⁶⁷

Katakana that show this clearly include ナ (top horizontal, then the descending vertical-with-hook), タ (the upper outer stroke first, then the inner stroke, then the central dot), マ (the curving body before the lower dot), ヤ (the upper short stroke first), and ラ (top horizontal before the long curving body).⁷⁸

The katakana parallels are visually crisper than their hiragana counterparts because the strokes are short and detached, with no continuous curve linking them.⁴

Left to right

大原則2 (左から右へ): write left elements before right elements, and within a single horizontal stroke, write left-to-right.⁶⁷

The clearest katakana demonstrations are イ (left short tick, then the long right vertical-with-curve), リ (left short stroke, then the longer right stroke), ハ (left diagonal before right), and ホ (left vertical first, then the central crossbar, then the right elements).⁷⁸

The katakana versions are visually crisper than the hiragana い/り/は/ほ because the right-side element is detached from the left, not connected by a curve.⁴

Horizontal before vertical when they cross

原則1 (横画が先): when a horizontal and a vertical stroke cross, write the horizontal first.⁶⁷

In katakana, this rule governs ナ (top horizontal, then the long descending vertical with hook), チ (top short horizontal, then the long crossing diagonal, then the descending body), テ (top short horizontal, then the longer crossing horizontal, then the descending stroke), and the stacked-horizontals-then-vertical pattern of モ.⁶⁷⁸

Katakana preserves the kanji-shape memory of these crossings more transparently than hiragana does, because the source kanji components (ナ from 奈, チ from 千, テ from 天, モ from 毛) are still recognizable in the kana shape.⁵⁹

The short-stroke "dot" and where it lands

A diagonal short stroke, the 点 (ten, "dot"), appears in イ, ウ, カ, ク, タ, ナ, and several others.⁸ It is the katakana-specific application of the directional principles: when a short tick sits to the right of or below a longer anchoring stroke, the longer stroke is written first; when the tick sits above-left, it is written first.⁶⁷⁸

In practice this means the tick on カ comes after the main left vertical-with-hook, while the upper-left tick on ヤ comes before the long curving body. The rule is one piece of the directional toolkit, not a separate convention.

Enclosing strokes are rare in katakana

原則6 / 原則7 (貫く縦画は最後 / 貫く横画は最後): a stroke that runs through or closes a character is written last.⁶ 原則4 (外側が先) governs full enclosures, where the outer frame is written before the inside but the closing bottom stroke is last.⁶⁷

In hiragana, the enclosure-last rule does heavy work: ぬ, ね, れ, わ, お, and む all close with a final loop.⁷ In modern katakana, the rule applies in essentially one place: ロ, where three strokes (left vertical, top-and-right wrap, closing bottom horizontal) follow the same sequence as the kanji 口.⁶⁷⁸

The obsolete katakana ヰ and ヱ (removed from the standard set by the 1946 orthographic reforms) had more substantive enclosures; in the modern 46-kana set, ロ is the cleanest demonstration of the enclosure-last principle.⁴¹⁰¹¹

The three stroke-end types: tome, hane, harai

A stroke-order diagram shows sequence and direction. It also shows how each stroke finishes. Three endings are named in the standard pedagogy:⁷

Term	Kana	Meaning	What the diagram shows	Katakana example
止め	tome	"stop"	The stroke ends in a firm halt; in brush calligraphy a small pool of ink marks the end.	The angular corner of エ; the inner corner of コ.
跳ね	hane	"hook" / "flick"	The stroke ends with an upward flick that diverges from the main direction.	The lower hook of ク, ワ, ル, レ; the inner hook on シ's sweep.
払い	harai	"sweep"	The pen lifts gradually while still in motion, tapering to a thin tail.	The diagonals of ノ, メ, ア, ハ, ス.

Compared with hiragana, where 払い sweeps end many kana (し, つ, て, の, ふ, the diagonals of な), katakana relies more heavily on 止め at angular corners.⁷⁴ The same three terms describe what the chart shows. In katakana, the balance shifts toward 止め because the script is angular rather than cursive.

筆順 and 書き順 are interchangeable

筆順 (hitsujun) is the formal term used by the MEXT guide and academic sources; 書き順 (kakijun) dominates everyday speech and search queries.⁶ Both refer to the same thing.

Stroke order for all 46 base katakana

How to read the chart

Each character is shown with numbered strokes (1, 2, 3, …) indicating the sequence. A dot or arrow marks the start point and direction of each stroke. The shape at the tail of each numbered stroke suggests the ending style (止め, 跳ね, or 払い).⁷ The orders below follow the MEXT 1958 standard as taught in Japanese elementary education and as encoded in the KanjiVG open dataset.¹²⁸

Print-versus-textbook differences are smaller in katakana than in hiragana, because the angular shapes leave less room for print fonts to fuse cursive strokes. Two cases are worth flagging up front. First, the ソ vs ン and シ vs ツ start-direction distinction is a stroke-order rule, not a font quirk.⁴ Second, the question of whether リ's two strokes touch is resolved by the textbook (教科書体) form keeping them detached.⁸

The five vowels: ア イ ウ エ オ

• ア (2 strokes): top short horizontal-into-diagonal (left-to-right, 払い at the tip), then the long descending hook on the right. Source kanji: 阿.⁹⁸¹²
• イ (2 strokes): left short diagonal (the "tick") first, then the long descending vertical-with-curve on the right. Source kanji: 伊.⁹⁸
• ウ (3 strokes): top short stroke (the dot), then the upper horizontal-with-hook, then the central descending stroke. Source kanji: 宇.⁹⁸
• エ (3 strokes): top horizontal, then the central vertical, then the bottom horizontal. The top-to-bottom ordering across all three positions demonstrates 大原則1 and 原則1 together. Source kanji: 江.⁹⁸
• オ (3 strokes): top horizontal, then the long vertical that crosses it (with the hook on the lower-left), then the short right-side diagonal. Source kanji: 於.⁹⁸

k-row: カ キ ク ケ コ

• カ (2 strokes): left vertical-with-hook (descending and curving outward at the bottom), then the right short diagonal. The shape preserves the left side of the source kanji 加.⁵⁹⁸
• キ (3 strokes): top diagonal-into-horizontal, then the lower horizontal, then the long descending vertical that crosses both. Source kanji: 幾.⁹⁸
• ク (2 strokes): top short diagonal stroke (払い), then the long curving body that descends to the lower-left. Source kanji: 久.⁹⁸
• ケ (3 strokes): top short diagonal, then the long descending vertical-with-hook on the left, then the short crossing diagonal on the right. Source kanji: 介.⁹⁸
• コ (2 strokes): top horizontal-into-right-angle (the upper-right corner stroke), then the bottom horizontal closing the shape. Source kanji: 己.⁹⁸

s-row: サ シ ス セ ソ

• サ (3 strokes): top horizontal, then the long descending vertical on the left, then the short crossing diagonal on the right. Source kanji: 散.⁹⁸¹³
• シ (3 strokes): two short diagonal strokes stacked on the upper-left (each drawn top-down), then a long sweeping stroke from lower-left up to the upper-right. The bottom-up direction of the third stroke is the canonical distinguisher from ツ. Source kanji: 之.⁷⁴⁹⁸
• ス (2 strokes): top short horizontal-into-diagonal (descending right-to-left), then the long crossing diagonal that descends left-to-right and ends with a 払い sweep. Source kanji: 須.⁹⁸
• セ (2 strokes): top horizontal-into-right-angle (the long curving stroke forming the upper and right portions), then the crossing horizontal on the left. Source kanji: 世.⁹⁸
• ソ (2 strokes): top short diagonal stroke (drawn top-to-bottom), then the long descending stroke that starts at the upper-right and sweeps down-left. The top-down direction of the second stroke is the canonical distinguisher from ン. Source kanji: 曽.⁷⁴⁹⁸

t-row: タ チ ツ テ ト

• タ (3 strokes): top outer diagonal (the upper-left-to-lower-right stroke forming the outer frame), then the inner crossing diagonal, then the central short stroke on the lower inside. Source kanji: 多.⁹⁸
• チ (3 strokes): top short diagonal, then the long crossing horizontal, then the long descending stroke that crosses through. Source kanji: 千.⁹⁸
• ツ (3 strokes): two short diagonal strokes stacked side-by-side on the upper portion (each drawn top-down, left then right), then a long sweeping stroke from upper-right curving down to the lower-left. All three strokes go top-down; this is the canonical distinguisher from シ. Source kanji: 川.⁷⁴⁹⁸
• テ (3 strokes): top short horizontal, then the longer crossing horizontal beneath it, then the long descending body. Source kanji: 天.⁹⁸
• ト (2 strokes): long descending vertical on the left, then the short diagonal on the upper right. Source kanji: 止.⁹⁸

n-row: ナ ニ ヌ ネ ノ

• ナ (2 strokes): top horizontal, then the long descending vertical-with-hook crossing through. Source kanji: 奈.⁹⁸
• ニ (2 strokes): top short horizontal, then the longer bottom horizontal beneath it. The simplest possible demonstration of 大原則1. Source kanji: 仁.⁹⁸
• ヌ (2 strokes): the upper outer stroke (the curving stroke forming the top and left), then the long descending diagonal that crosses through and ends with a 払い sweep on the lower-right. Two strokes total, no enclosed loop. Source kanji: 奴.⁹⁸
• ネ (4 strokes): top short stroke (the upper dot), then the horizontal beneath it, then the long descending vertical, then the short angled finish on the lower-right. The four-stroke count makes ネ one of the densest katakana by stroke count. Source kanji: 祢.⁹⁸
• ノ (1 stroke): a single descending diagonal from upper-right to lower-left, ending in a 払い sweep. Source kanji: 乃.⁷⁹⁸

h-row: ハ ヒ フ ヘ ホ

• ハ (2 strokes): left short diagonal (払い at the tip), then the right short diagonal (also 払い). The two strokes do not touch in the textbook form. Source kanji: 八.⁹⁸
• ヒ (2 strokes): top short diagonal (the upper-left stroke), then the long horizontal-into-hook forming the lower body. Source kanji: 比.⁹⁸
• フ (1 stroke): a single stroke that begins as a short horizontal at the top, turns at the upper-right corner, and descends as a diagonal sweep ending in a 払い. Source kanji: 不.⁷⁹⁸
• ヘ (1 stroke): a single angled stroke, left-to-right with a peak in the middle. The glyph is essentially identical to hiragana へ, and both are one stroke. Source kanji: 部.⁷⁹⁸
• ホ (4 strokes): top short horizontal, then the long descending vertical, then the left diagonal, then the right diagonal. The four short detached strokes preserve the structure of the source kanji 保.⁹⁸

m-row: マ ミ ム メ モ

• マ (2 strokes): the curving body stroke (top horizontal turning into a descending diagonal sweep on the lower-left), then the short stroke on the lower inside. Source kanji: 末.⁹⁸
• ミ (3 strokes): three short diagonal strokes stacked top to bottom, each drawn left-to-right with a slight downward slope; the bottom stroke is slightly longer. The cleanest katakana demonstration of 大原則1. Source kanji: 三.⁹⁸
• ム (2 strokes): short stroke on the upper-left (a small diagonal "tick"), then the long curving body stroke forming the lower portion and the right side. Source kanji: 牟.⁹⁸
• メ (2 strokes): the descending diagonal from upper-right to lower-left, then the crossing diagonal from upper-left to lower-right (払い at the tip). Source kanji: 女.⁹⁸
• モ (3 strokes): top horizontal, then the lower horizontal beneath it, then the long descending vertical-with-hook crossing through. The vertical comes last because both horizontals must precede it: 大原則1 and 原則1 stacked. Source kanji: 毛.⁶⁷⁹⁸

y-row: ヤ ユ ヨ

• ヤ (2 strokes): short stroke on the upper-left (the small "tick"), then the long descending stroke that curves through and ends with a 払い. Source kanji: 也.⁹⁸
• ユ (2 strokes): the upper horizontal-into-right-angle (forming the top and right), then the bottom horizontal closing the shape. Source kanji: 由.⁵⁹⁸
• ヨ (3 strokes): top horizontal-into-right-angle (the upper-right corner stroke), then the middle horizontal that crosses inside, then the bottom horizontal closing the shape. Three horizontals descending top-to-bottom with the right side joining them. Source kanji: 与.⁹⁸

r-row: ラ リ ル レ ロ

• ラ (2 strokes): top short horizontal, then the long curving body that descends and sweeps left. Source kanji: 良.⁹⁸
• リ (2 strokes in the textbook form): left short stroke (top-to-bottom), then the longer right stroke (top-to-bottom, sometimes with a slight hook at the bottom). The two strokes do not touch in the textbook (教科書体) form. Source kanji: 利.⁹⁸
• ル (2 strokes): left short curving stroke, then the long descending stroke on the right ending with a 跳ね hook on the upper-right. Source kanji: 流.⁹⁸
• レ (1 stroke): a single stroke beginning as a short descending diagonal on the upper-left, turning at the lower corner, and sweeping up to the upper-right with a 跳ね hook. Source kanji: 礼.⁷⁹⁸
• ロ (3 strokes): left vertical, then the upper-and-right wrap (horizontal-into-right-angle forming the top and right side), then the bottom horizontal closing the shape. The cleanest enclosure-last demonstration in modern katakana, mirroring the stroke order of the kanji 口. Source kanji: 呂.⁶⁷⁹⁸

w-row and n: ワ ヲ ン

• ワ (2 strokes): top horizontal-into-right-angle (the upper stroke forming the top and the right side), then the long descending stroke ending with a 跳ね hook. The shape is identical to ウ minus the top dot. Source kanji: 和.⁹⁸
• ヲ (3 strokes): top horizontal, then the middle horizontal, then the long descending crossing stroke ending with a 跳ね hook. Source kanji: 乎.⁹⁸
• ン (2 strokes): a small diagonal "tick" on the upper-left, then a long sweeping stroke from lower-left curving up to the upper-right. The bottom-up direction of the second stroke is the canonical distinguisher from ソ. Source kanji: 尓.⁴⁵⁹⁸

The 46-kana count is the post-1946 teaching set

The "46 base katakana" are the standard set used in modern Japanese education. The 1946 Cabinet notification 現代かなづかい deemed ヰ (wi, from 井) and ヱ (we, from 恵) obsolete, replacing them with イ and エ in all standard contexts.¹⁰¹¹ All stroke orders in this section describe the post-1946 standard 46-kana set as taught in Japanese elementary education and encoded in the KanjiVG dataset.¹²³⁸

Why katakana is generally simpler stroke-wise than hiragana

Katakana inherits from kanji components, not from cursive simplification

Each katakana is one component lifted from a man'yōgana kanji. The 片 in 片仮名 literally means "fragment" or "partial", a direct reference to the lifted-component origin.⁴⁵ By contrast, each hiragana is the full cursive (草書) trace of a kanji, retaining the entire shape in simplified form.¹⁴

The lifted-component origin produces shorter, straighter shapes. カ is the left side of 加; キ derives from the upper portion of 幾; テ is taken from 天; ホ from 保.⁵⁹ In each case the katakana is recognizable as one part of its source kanji, not as a cursive whole.

Katakana was developed in the 9th century during the early Heian period by Buddhist monks in Nara. They took parts of man'yōgana characters as a form of shorthand to mark pronunciation guides (送り仮名 and 訓点) on Chinese-language Buddhist texts.⁴⁵¹⁴

Stroke counts and contours run lighter on average

The summary is not "katakana is always fewer strokes." It is "katakana shapes are mechanically lighter per stroke," because each stroke tends to be short, straight, and detached rather than long, curving, and connecting.⁴⁸ Comparing pairs across the 46-kana set:

Hiragana	Strokes	Katakana	Strokes	Notes
ふ	4	フ	1	Single-stroke katakana, four-stroke hiragana
ぬ	2	ヌ	2	Same count; ぬ has an enclosed loop, ヌ has none
ま	3	マ	2	Lighter in katakana
ね	2	ネ	4	One of the few cases where katakana is heavier
ほ	4	ホ	4	Tie on count; ホ is four detached strokes
お	3	オ	3	Tie on count; オ has no closing loop
し	1	シ	3	Katakana heavier; the three short strokes separate
つ	1	ツ	3	Same pattern as し vs シ

The defensible claim is qualitative: katakana strokes are shorter, straighter, and more often detached, even when the per-kana count is similar.

Angular corners replace curves

Katakana favors straight segments that meet at sharp angles (止め endings at corners). Hiragana favors continuous curves with 払い sweeps.⁷⁴ The principle "horizontal goes left-to-right, vertical goes top-to-bottom" is more visible in katakana because the strokes are short and detached.

A learner can often read stroke order from the shape alone in katakana. This is harder in hiragana, where curve-connections obscure the underlying directional rule.⁶⁷ This visual cue lets a beginner self-correct stroke order without consulting a chart for each character. Short straight strokes commit to one direction each, and the directions stack predictably under the 大原則.

Fewer enclosed loops to close last

The enclosure-last principle dominates hiragana stroke order: ぬ, ね, れ, わ, お, and む all end with a closing loop, and several others (ま, ゆ, よ) contain enclosed segments mid-character.⁶⁷⁸

Modern katakana has essentially one closing-stroke kana: ロ, where the bottom horizontal lands last, mirroring the kanji 口.⁶⁷⁸ The obsolete ヰ and ヱ had more substantial enclosures but sit outside the modern teaching set.¹⁰¹¹ With fewer cases like this, learners have less to track about where the closing stroke goes. That is part of why katakana feels mechanically simpler to a learner who has already drilled hiragana.⁴

Stroke direction does most of the disambiguation work

In hiragana, closing-loop direction distinguishes ぬ from め and shapes the family of わ, ね, and れ.⁷ In katakana, stroke direction does the disambiguation work. The canonical demonstrations are the シ/ツ and ソ/ン pairs.

Getting stroke order right therefore matters more for reading handwritten katakana, not less.⁴ The deeper treatment of all confusable pairs belongs in a dedicated lookalikes article; this article previews the シ/ツ/ソ/ン case as the canonical demonstration.

When you can de-prioritize katakana stroke order

The fair version of the question is the same as in hiragana: if a learner will only ever type Japanese, is the time spent on per-katakana stroke order recoverable? The honest answer is mostly yes. There are two specific exceptions and one disambiguation drill that pays back regardless.

Typing-only learners can defer per-kana drill

A learner whose Japanese goals are typing-only (chat, reading on a phone, IME input from a romaji keyboard) can defer per-katakana stroke-order mastery without losing access to the language.¹⁵ Recognition reading does not require handwriting each katakana, and romaji-based IME input bypasses the handwriting recognizer entirely.¹⁶¹⁷

The cost falls due on kanji and on phone handwriting input

Two specific points where the deferred cost comes due:

1. Kanji study. Kanji stroke order is the same system as kana stroke order, multiplied across thousands of characters. The MEXT guide's principles were defined for the 881 educational kanji first and then declared to apply equally to kana and Roman letters.¹² A learner who skips the principles for 46 katakana usually has to re-learn them, more painfully, with the first few hundred kanji.²⁶³
2. Handwriting input on phones and tablets. Zinnia, a widely used open-source online handwriting recognizer, processes user input as a time-ordered "sequence of coordinate data" from the pen strokes.¹⁶ Production handwriting lookup tools instruct users to draw in correct stroke order for accurate results,¹⁷ and academic real-time evaluators sample pen-trajectory features per stroke and compare them against the correct order to score learner writing.¹⁸ KanjiVG, the open dataset behind several recognizers and stroke-animation displays, stores each katakana as an ordered sequence of stroke paths. The data model itself presumes a canonical stroke order.⁸

Lookalike disambiguation buys most of the value

If a learner drills any per-katakana stroke order, the シ/ツ/ソ/ン set is where it pays the largest reading dividend.⁴ The stroke-direction distinction separates the four glyphs in handwriting. Reading handwritten or stylized katakana depends on parsing that distinction correctly. The deeper treatment of all confusable katakana pairs belongs in a dedicated lookalikes article; this article previews the シ/ツ/ソ/ン case as the canonical demonstration.

Speed of acquisition is not the same as long-term cost

Skipping katakana stroke order shortens week one of katakana study. It does not shorten the time spent learning the first few hundred kanji, where the same principles return at scale.¹²³

Good to know

シ/ツ and ソ/ン start direction is a stroke-order rule, not a font quirk

The シ/ツ and ソ/ン pairs look like mirror-image siblings in print. In handwriting, the direction of the long stroke is what distinguishes them.⁴ シ's third stroke starts at the lower-left and sweeps up to the upper-right (bottom-up); ツ's third stroke starts at the upper-right and sweeps down to the lower-left (top-down). ソ's second stroke goes top-down; ン's second stroke goes bottom-up.⁷⁴⁸ Drawing シ with a top-down final stroke produces ツ, not a slightly wrong シ.

ヲ is taught but rarely handwritten by adults

The particle を is written in hiragana in modern Japanese. The katakana ヲ appears mostly in pre-WW2 official documents (which used kyūjitai, older character forms, plus katakana as the default mix), in stylized typography, and in ruby annotations (small pronunciation guides).⁴ The 1946 orthographic reforms flipped the default for official prose to hiragana plus kanji, leaving ヲ functionally orphaned in everyday handwriting.⁴¹⁰¹¹ Recognizing ヲ matters; writing it routinely does not.

The chōonpu ー is one horizontal stroke, drawn left-to-right

The chōonpu ー is a single horizontal line written in the center of the text line, with the width of one kana character. It is written horizontally in horizontal text and vertically in vertical text.¹⁹ It is used primarily with katakana to mark a long vowel in loanwords (タクシー, コーヒー, ラーメン). It appears rarely in hiragana and not at all in standard romanization.¹⁹ One stroke, left-to-right, finished with a clean 止め.⁷¹⁹

The katakana repetition marks ヽ and ヾ are largely confined to proper names

ヽ (unvoiced) repeats the preceding katakana syllable, and ヾ (voiced, with dakuten, the two-dot voicing mark) repeats it with added voicing.²⁰ Both are "generally not used in modern Japanese outside proper names, though they may appear in informal handwritten texts."²⁰ Each is drawn as a single short diagonal stroke. The dakuten on ヾ is added as a separate two-dot mark in the upper-right, following the base-character-first rule.⁶²⁰ A beginner needs to recognize them on signage and in older names but rarely has occasion to write them.

Print-vs-textbook divergence is smaller in katakana than in hiragana

The clearest case is リ, which in print fonts sometimes appears with the two strokes touching at the top. The textbook (教科書体) and handwritten form keeps them detached.⁸ Some print fonts also flatten the slant differences in シ/ツ and ソ/ン, so the four glyphs look more alike than they do in handwriting. The canonical handwritten forms preserve the stroke-direction distinction visibly.⁴⁸ The print form is what your fonts show. The handwritten form is what your phone IME, your kanji recognizer, and a Japanese reader's eye expect.¹⁷¹⁸

The stroke orders feel like kanji because they are kanji components

Every katakana derives from one component of a specific man'yōgana kanji, lifted and standardized. The standard derivations are: ア from 阿, イ from 伊, ウ from 宇, エ from 江, オ from 於, カ from 加, キ from 幾, ク from 久, ケ from 介, コ from 己, サ from 散, シ from 之, ス from 須, セ from 世, ソ from 曽, タ from 多, チ from 千, ツ from 川, テ from 天, ト from 止, ナ from 奈, ニ from 仁, ヌ from 奴, ネ from 祢, ノ from 乃, ハ from 八, ヒ from 比, フ from 不, ヘ from 部, ホ from 保, マ from 末, ミ from 三, ム from 牟, メ from 女, モ from 毛, ヤ from 也, ユ from 由, ヨ from 与, ラ from 良, リ from 利, ル from 流, レ from 礼, ロ from 呂, ワ from 和, ヲ from 乎, ン from 尓.^4591213 The obsolete pair is ヰ from 井 and ヱ from 恵. Both were deemed obsolete in 1946 but are still used stylistically in brand names like ニッカウヰスキー and ヱビス.¹⁰¹¹

Knowing this collapses "katakana stroke-order memorization" into "draw the lifted component of a kanji you already know or will soon learn."⁴⁵

Common beginner anti-patterns

Three patterns appear again and again in first-week katakana drills:

• Drawing シ bottom-up in the wrong slot, or all of ツ's strokes the wrong way. The シ/ツ direction rule is binary: シ's long stroke is bottom-up, ツ's three strokes are all top-down.⁷⁴⁸ Mirroring the rule flips the kana.
• Joining the two strokes of ハ into one continuous shape. ハ inherits its shape directly from the source kanji 八, which is itself two detached strokes; the katakana preserves that detachment, and a single curving stroke produces something closer to cursive 八 than to ハ.⁹⁸
• Drawing four-stroke kana (ホ, ネ) as a single continuous contour. The canonical stroke counts (ホ = 4, ネ = 4) define where the 止め, 跳ね, and 払い endings land. Collapsing the strokes erases those endings and degrades the shape for both human readers and handwriting recognizers.⁷⁸¹⁷¹⁸

See also

• Katakana Mnemonics That Actually Work
• Hiragana vs. Katakana: How to Tell Them Apart and Use Both
• How to Write Your Name in Katakana: Foreign-Name Transcription Rules with Examples
• How to Practice Writing Hiragana: A Drill Plan, Free Sheets, and the Anki Hybrid
• Writing Kanji by Hand: Is It Still Worth It?
• Hiragana, Katakana, or Kanji First? A Beginner's Script Order

References

Footnotes

1. 文部省. 『筆順指導の手びき』. 博文堂, 1958 (昭和33年). National Diet Library record: https://ndlsearch.ndl.go.jp/en/books/R100000002-I000000984674 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

2. 「筆順指導の手びき」(「本書のねらい」等を引用). 漢字の正しい書き順(筆順). https://kakijun.jp/main/hitusjunsidonotebiki.html ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³

3. 文部科学省. 『小学校学習指導要領』第2章 各教科 第1節 国語. https://www.mext.go.jp/a_menu/shotou/cs/1319951.htm ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

4. Katakana. Wikipedia. https://en.wikipedia.org/wiki/Katakana ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶ ↩²⁷

5. 片仮名. ウィキペディア日本語版. https://ja.wikipedia.org/wiki/%E7%89%87%E4%BB%AE%E5%90%8D ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹

6. 筆順. ウィキペディア日本語版. https://ja.wikipedia.org/wiki/%E7%AD%86%E9%A0%86 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²²

7. Stroke order. Wikipedia. https://en.wikipedia.org/wiki/Stroke_order ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶ ↩²⁷ ↩²⁸ ↩²⁹ ↩³⁰ ↩³¹ ↩³² ↩³³

8. KanjiVG project. http://kanjivg.tagaini.net/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶ ↩²⁷ ↩²⁸ ↩²⁹ ↩³⁰ ↩³¹ ↩³² ↩³³ ↩³⁴ ↩³⁵ ↩³⁶ ↩³⁷ ↩³⁸ ↩³⁹ ↩⁴⁰ ↩⁴¹ ↩⁴² ↩⁴³ ↩⁴⁴ ↩⁴⁵ ↩⁴⁶ ↩⁴⁷ ↩⁴⁸ ↩⁴⁹ ↩⁵⁰ ↩⁵¹ ↩⁵² ↩⁵³ ↩⁵⁴ ↩⁵⁵ ↩⁵⁶ ↩⁵⁷ ↩⁵⁸ ↩⁵⁹ ↩⁶⁰ ↩⁶¹ ↩⁶² ↩⁶³ ↩⁶⁴ ↩⁶⁵ ↩⁶⁶

9. "Appendix: Comparison of hiragana and katakana derivations." Wiktionary. https://en.wiktionary.org/wiki/Appendix:Comparison_of_hiragana_and_katakana_derivations ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶ ↩²⁷ ↩²⁸ ↩²⁹ ↩³⁰ ↩³¹ ↩³² ↩³³ ↩³⁴ ↩³⁵ ↩³⁶ ↩³⁷ ↩³⁸ ↩³⁹ ↩⁴⁰ ↩⁴¹ ↩⁴² ↩⁴³ ↩⁴⁴ ↩⁴⁵ ↩⁴⁶ ↩⁴⁷ ↩⁴⁸ ↩⁴⁹ ↩⁵⁰

10. Wi (kana). Wikipedia. https://en.wikipedia.org/wiki/Wi_(kana) ↩ ↩² ↩³ ↩⁴ ↩⁵

11. We (kana). Wikipedia. https://en.wikipedia.org/wiki/We_(kana) ↩ ↩² ↩³ ↩⁴ ↩⁵

12. A (kana). Wikipedia. https://en.wikipedia.org/wiki/A_(kana) ↩ ↩²

13. Sa (kana). Wikipedia. https://en.wikipedia.org/wiki/Sa_(kana) ↩ ↩²

14. How did katakana and hiragana originate? sljfaq.org. https://www.sljfaq.org/afaq/originofkana.html ↩ ↩²

15. Japan Foundation and Japan Educational Exchanges and Services. "JLPT Levels Summary." https://www.jlpt.jp/e/about/levelsummary.html ↩

16. Kudo, Taku. "Zinnia: Online hand writing recognition system with machine learning." Project page. https://taku910.github.io/zinnia/ ↩ ↩²

17. Kanji recognition system documentation. sljfaq.org. https://kanji.sljfaq.org/ and https://www.sljfaq.org/afaq/stroke-order.html ↩ ↩² ↩³ ↩⁴

18. Hatano, Shōji et al. "Robust and real-time stroke order evaluation using incremental stroke context for learners to write Kanji characters correctly." Pattern Recognition Letters, vol. 117, 2019, pp. 9–16. https://www.sciencedirect.com/science/article/abs/pii/S0167865518303258 ↩ ↩² ↩³

19. Chōonpu. Wikipedia. https://en.wikipedia.org/wiki/Ch%C5%8Donpu ↩ ↩² ↩³

20. Iteration mark. Wikipedia. https://en.wikipedia.org/wiki/Iteration_mark ↩ ↩² ↩³