Scholastic sunflower.

In the first Jewish school built in Berlin since the Nazi times, Zvi Hecker has used his obsession with geometry to generate a network of memorable particular places to act as a humane backdrop to education.

The borders of Wilmersdorf and Grunewald offer that calm, leafy background of loose-fit into which villas, highways, a bombastic Olympic Stadium or a metal-sheathed congress hall can be dropped without disturbance: only just an identifiable context, if you exclude the trees. The site of Hecker's last major building, the Spiral in Ramat Gan, on the edge of Tel Aviv (AR October 1990) might be similarly described as an absorptive condition in which to sit a powerful gyrating body.

Both the earlier building and the new school have a gyratory system as an a priori. The Israeli building is jagged, wild, deliberately attracting a surface of bricolage and punctuated by sharp metal slashes. The German building rests the gyrations on the ground and articulates the flanks with a controlled system: all straight walls are white, all curved walls are grey, all 'snakes' are metal, all evacuated spaces can be walked on or planted in. The metal slashes are here too, but as a controlled reference to the system of organisation, accompanied by other frame elements. The openings in the building are calm, firm and consistent.

Yet the new building is in every way the more extraordinary of the two. It is full of incident and exploitation of a local occurrence, full of nooks and crannies, full of subtle variations of shape and size: yet it is highly controlled. The 'Sunflower' which is the generating idea is always traceable but not overbearing.

The clue lies in Hecker's process of working. He makes a continual series of drawings: endlessly, right up to the last minute of detail construction. He 'sits' as much as possible on the site. He takes fully-worked and dimensioned drawings and still scribbles away over them, modifying - nay - honing them: rarely satisfied, yet, in this building, too intelligent to proliferate elements but too ambitious to standardise spaces. Such a procedure (which may derive from the spiritual mannerism of Israelis to endlessly discuss and negotiate) was bound to be difficult in a culture that thrives upon prearranged procedures. The role of Inken Bailer (who also guided ourselves and Herman Hertzberger so deftly through the IBA minefields: AR April 1987) was critical. She somehow devised a procedural method in which Zvi could track over the layout, caress the apertures and hone the surfaces while the builders and bureaucrats could feel assured that the building would be on time and on budget.

Yet Hacker himself was clearly affected by both process and culture. He continues now to work in Berlin and more and more assertively points to the influences from Europe upon his work. He was already a second year student at Cracow when his family emigrated to Israel. Moreover, he became the protege and eventual partner of Alfred Neumann who had himself worked for Behrens and Perret. The process of defining the trajectory and rationalising the elements - as well as continuously having to walk members of your peer group (whether Gehry, Hejduk or Libeskind) around your building site puts on a special pressure. Almost like being at architecture school, with a series of ongoing crits. A far cry from the Israeli scene where jibes of 'eccentric' and 'show-off' serve as the critical base.

A conscious evaluation of light and the role of the window emerges - less as ventilator more as picture-plane. Of the balcony as exotic and precious transition from the winter chamber into the spring breezes. Of shelter as necessity rather than form. Essentially Northern issues. And in parallel, an inevitable response to the construction process in a country where labour is expensive and material is cheap, time is money and dimensions are precise. Hacker becomes procedurally Northern (if his building is to happen) but retains his character as long as he can hone on. Thus the jagged elements become far more deliberate and sharp-edged than at Ramat Gan. The interior spaces far more celebratory.

Yet in the process of walking around the building, the Mediterranean experience is remembered. The left-hand side (seen from the street) is a knife cut through the sunflower system. Exposing a series of re-entrants and crevices, pieces jutting viciously and delving insidiously. As you move down, this very rich offering suggests the idea of density and a sure expectancy of much going on inside. Turning round into the rear playground the sheer range of the parts and the 'knitted' quality of the whole add to this. Of course, it is a town. What else could it be? And the total system reinforces the analogy. The radiating sweeps define 'quartiers' and their streets, the 'snakes', are a counter-movement, somewhat like a stream, the edges of the town have different physiognomies dependent upon circumstance: one tight, one heroic, one secret and one casually falling away.

The quality of external space is of a series of localities. Hecker has exploited this internally, sometimes placing windows so that at first, you only see a wash of light on a nearby wall and you have to move into them in order to turn towards a longer view. He is particularly keen on reflected light rather than direct light. In other cases, windows are the product of the many balconies that crop up. if this all sounds too tricky, you must remember that it is a large building with a very small number of window types. All staircases similar. Roof lights similar. Corridors similar in width (though when they are 'snakes' liable to undulate in both directions). The game is that of tweaking these repetitive elements into new juxtapositions.

In his own words, Hecker wants the school to be a 'big family house' rather than an institution. So there are several places in which he enjoys (and encourages) the fact that the kids can hide. He enjoys the fact that only the inmates really know all the routes through the building. Town rather than house. Though the outside spaces are often of a similar dimension to those inside and the progression from small nook to unfolding piece of Wilmersdorf Arcadia is by stealth. In England, the memory might be for the collegiate progression of space, but for Hecker, Cracow returns into the conversation. The Mediterranean characteristic pervades in the question of surface and incision. In only one part of the building does he (quite deliberately) offer a 'standard' two-storey run of repeated rooms and window-and-spandrel architecture. Nearby, though, is one of his naughtiest moments when a pathway disappears into the ground to crawl under a low-lying 'snake' bridge, accompanied alongside by a strip of water. Elsewhere, he has been at great pains to clarify the purity of the wall-window definition. He has borrowed a detail from his nineteenth-century Berlin office where the outer (timber) frame is visible from within the room but sheathed by a flange of outer wall. It enables the extremely efficient but 'chunky' German windows (with which we all have aesthetic problems) to do their job without over scoring the facade.

In many ways, this architecture defies categorisation. It is often discussed in the light of Hecker's earlier preoccupation with geometry and plan figuration. Yet in this school, it is more of a regulator of position than anything else. It is clearly Modernist in terms of the articulation of solids and exposure of faces, yet its spatial instincts are 2000 years old. It is European, in the care of its detail and the attention to indeterminate light, yet it is outgoing and still waiting for the hot sun and the need to crouch in the shade when you stand outside. It is aware of symbolism and the need to react to a bourgeois world (especially noticeable in the main assembly hall and on the internal handrails where the light wood facings tend to be acceptable to parents but diffuse the power of the architecture). Fortunately, it is not too careful.

So Hecker stands at a watershed. His Army Museum in Tel Aviv (also won in competition at roughly the same time as this building, but on a slower building programme), pulls away from gyrating geometry completely and starts to involve slow ramping as a device. It will be interesting to see how Tel Aviv can absorb the calmness of the North.

Hecker is always immersed in the work and is always conscious of cause-and-effect. The generation of one building out of another.

It will be interesting to listen to the (inevitable) comparisons that will be made between the school and Libeskind's Jewish Museum. The latter is surely much more of a citadel than this building. Yet symbolism and light preoccupy both.

The clue for the school lies in the legacy of Hecker's honing and taste for making. A tiny clue is his offer to expose a piece of blockwork or unfinished structure in a corner or two: like a picture, or a living wall-chart for the kids. 'Come and enjoy the stuff of the building, with me', he seems to say.

Surely an architect's interpretation of 'Beth-Sepher' . . . the house of the book . . . which is its basic form, by the way.