Hojicha is made by putting heat into tea leaf. Two questions follow from that. What happens inside the leaf, and what do 'light', 'medium' and 'deep' roast actually mean on a label? This article keeps to what published research and company descriptions state, without adding numbers or compound names of our own.

Heat creates the raw material of aroma

A study dealing with the roasting of bōcha explains that heating generates volatile compounds from amino acids and sugars. Components already present in the leaf take on heat and turn into different substances; the raw material of aroma is newly made at that point. Before and after the heat, the contents have been exchanged. Nothing was added — something changed.

So roasting is not a matter of adding aroma to tea. It raises aroma out of materials the leaf already holds, which is why the same leaf gives a different result when the fire changes. That is where this study's explanation stops. It is also enough to explain why the craft is interesting: how the heat is applied comes straight through into the cup.

It is not only toastiness that increases

NARO, the National Agriculture and Food Research Organization, built an aroma profile in a trial on tea finishing processes in order to look at qualitative change in aroma from roasting. The components said to have become more influential are described with smell words: sweet, fruity, green, roasty, spicy, earthy, floral and meaty. What makes that list worth reading is that the aromas are named in ordinary language — these are categories of perception, not technical indices.

Read the list and toastiness turns out to be only part of it. Sweet notes, floral notes and fruit-like notes are in there too. The same organisation also writes that roasting increases the characteristic toastiness and fullness and eases the grassy note. Roasting starts to look less like one flavour and more like a way of widening the range of aromas available.

It is not written, however, that every aroma compound increases uniformly with roasting. NARO also notes that the strength of tea roasting differs depending on the consuming region, and that the toastiness differs with it. What the word 'profile' points at is not more or less, but a change in how the aroma is composed. Which components come forward depends on the conditions, and quantity and composition are separate questions.

Light, medium, deep — and whose words they are

Ryuhoen describes light-roast hojicha as making use of green tea's umami, so that the toasty aroma arrives together with that umami and reads as refined. Deep-roast hojicha, it says, is roasted strongly, and its character is a toastiness through which a faint sweetness drifts. Note that toastiness appears on both sides: light against deep is not about whether toastiness is present, but about what kind of toastiness it is.

Maruyasu Chagyo sets out three levels across a line of powders with the roast degrees aligned. The light roast is bō hojicha — the stem part — with a toasty aroma, a natural sweetness and a fresh taste with little astringency. The medium combines stem and leaf, with body, umami and sweetness in balance. The deep roasts the leaf twice, deliberately scorching it, for a chocolate-like aroma, a heavier body and a mellow finish.

What stands out there is that the raw material changes along with the roast: stem, then stem and leaf, then leaf. Fire and material are moving at the same time, so this is not a comparison in which only the roast level was varied. It is worth counting that in when you read the three side by side.

No shared scale, and no shortcut to health claims

Neither source gives a nationwide temperature or time that separates light, medium and deep. Each company is describing how it finished its own product. The same word on two shelves does not guarantee the same thing inside the pack, so it is safer to assume that a common measure has not yet been established. Treat the words as a rough guide, then look at the raw material and at how the description is written.

One note on scope. These findings hold within the conditions of each trial, and because the NARO trial addressed finishing processes, they cannot be stretched to cover commercial hojicha manufacturing conditions or the aroma change of every product on sale. More aroma compounds is also a separate matter from being good for you. This article goes as far as the doorway of how aroma is made, and no further; beyond that, the quickest route is to read the cited research itself.