Diary: camping in Wolverley, Kidderminster

We were recently camping for a few days at Wolverley, just north of Kidderminster (about 20 miles west of Birmingham). The proximate reason was to move our base c. 100 miles north and see some new National Trust sites.

My Honor 10 AI phone in its 'art' mode: our tent

Hanbury Hall NT: an apparently 2D monument

The Queen's Head, Wolverley where we had dinner

We left at eight as the band (above) was setting up. They were scheduled to start at nine. I said to Clare that there are thousands of perfectly competent bands like this up and down the country - even many about our age. The difference between them and the top acts is one of energy and charisma, the sheer animal spirits of Liam-style swagger.

I can think of many less attractive occupations than playing in a gigging pub band. Alas, my guitar skills were always bounded by my clumsiness. Perhaps in another life.

Rock’n’Roll Guns for Hire: the Sideman

BBC4's "Rock’n’Roll Guns for Hire: The Story of the Sideman " was a ninety minutes look at those popular music artists who support the star - but aren't members of the band.

Bernard Fowler and Earl Slick

Here is what The Guardian had to say:

"Earl Slick played guitar on stage for David Bowie, on and off for 40 years, so he has some insight into the psychology of being a sideman – a professional musician in the service of a big ego. “Most of the time we’re invisible,” he says. “Ghosts at the top table.”

Slick took an unusual leading role in Rock’n’Roll Guns for Hire: The Story of the Sideman (BBC4), exploring what it takes – and what it means – to have a career based on facilitating someone else’s vision. Even in this he was overshadowed, as big names offered their perfect sideman’s job description. “The better you are at your job, the less people will notice you,” said Keith Richards. “And that’s the whole point.”

In fact, all of the Rolling Stones turned out to heap measured praise on Bernard Fowler, their long-serving back-up singer, arranger and person in charge of remembering how all the songs go. The Stones seem utterly reliant on him.

We were also introduced to Wendy and Lisa of Prince and the Revolution fame; Crystal Taliefero, who played just about everything for Billy Joel; and legendary Stax guitarist Steve Cropper, whose co-writing credits on a string of hits make him a rare creature – a sideman with a pension."

What made this programme riveting was the psychological angle. The 'sidemen' seemed to be more professional, to have greater musical knowledge and certainly better musicianship than the stars they supported.

What on earth did they lack?

In a word: charisma. This was shown most clearly at the end, where after the death of Bowie some of the featured sidemen joined up to create a kind of 'Bowie tribute band'.  Bernard Fowler did the vocals - in his own accomplished and unique style - while Earl Slick played guitar. The whole thing was an embarrassing travesty - about as much excitement as watching paint dry.

"Low energy," as Trump might observe.

---

We know something about charisma. On the five-factor model it loads on Extraversion and Neuroticism (negatively). Considerably greater insight comes about through delving deeper into the 30 constituent facets (each of the five FFM dimensions further resolves into six facets).

Jasmine Vergauwe, Bart Wille, Joeri Hofmans and Filip De Fruyt wrote "Development of a Five-Factor Model charisma compound and its relations to career outcomes", concluding:

"In summary, the experts described the prototypical charismatic leader to be low on several Neuroticism facets, indicating that that they are in general:

• relaxed, unconcerned, cool: (low on anxiety) 
• optimistic (low on depression) 
• self-assured, glib, shameless (low on self-consciousness)
• clear-thinking, fearless, and unflappable (low on vulnerability).

Moreover, the experts rated the charismatic leader as typically high on all Extraversion facets, except for excitement seeking. This means that the charismatic leader tends to be:

• cordial, affectionate, attached (high on warmth) 
• sociable, outgoing (high on gregariousness)
• dominant, forceful (high on assertiveness)
• vigorous, energetic, active (high on activity)
• happy, cheerful, and joyous (high on positive emotions). 

Further, two Openness facets have been indicated to be prototypically high for the charismatic leader, namely:

• actions (unconventional, eccentric)
• values (permissive, broad-minded).

Finally, within the Conscientiousness domain:

• achievement striving (workaholic, ambitious) 

is perceived to be high in charismatic leaders, and none of the Agreeableness facets came out as a relevant personality-related description of the prototypical charismatic leader."

The sidemen by contrast were introverted, transactional, very amiable and conscientious.

It occurred to me suddenly that they were the quintessential consultants.

---

In sport too the same dichotomy. It's a problem when the qualities that make for sporting success are not aligned with a native charisma:

• Djokovic has charisma, Murray is a sideman
• Peter Sagan and Alberto Contador have charisma; Froome, Quintana - sidemen
• Lewis Hamilton has charisma, Nigel Mansell was a sideman.

Trees

A. A. Gill used to say that the little secret of TV was that it was mostly watched by women.

I assume he made an exception for the footie.

The feminisation of everything - "How did it make you feel?" - is evident in factual programs. The Sky at Night is a lost cause, but I was pleased to watch the BBC4 program, "Sword, Musket & Machine Gun", which seemed to be bloke TV.

Arguably, "How did it make you feel?" is rather superfluous.

At any rate, I was watching it avidly while Clare seemed glued to her tablet for once.

---

Weights this morning, lowered saddle on exercise bike (better!), skipped lunch. The Christmas effect has not yet disappeared according to my mirror. Then walked to Morrisons w. Clare to buy new mop bucket.

Waiting for gales and predicted weekend snow.

---

Another Christmas effect - the 2016 Waitrose Christmas Tree has become the 2017 garden tree.

As a bonus, its tub will make a neat wastebin, once the mud's been washed off.

---

Continuing on the subject of trees, my hobbyist AI project continues. I'm writing tree-search functions (bounded depth-first) and continuing to learn oh-so-many Common Lisp features.

The idea is to define a number of different kinds of game-players: random, human-powered, depth-first to some depth, complete generation of game tree + minimax .. for 'Noughts and Crosses'.

Once you have the (rather extensive) apparatus assembled, the tools can be generalised to more ambitious applications.

It has reminded me of the difficulties of programming: large, branching trees are huge sprawling objects, hard to visualise, and recursive operations often intricate. As always, subtle (and not-so-subtle) bugs dominate the ecosystem.

---

Here's my example tree from Wikipedia:

A node (below) is a triple comprising: node name, followed by question mark (may later become the best move) followed by the leaf non-zero score for testing minimax.

((N0 ? 0) (((N11 ? 0) (((N21 ? 0) (((N31 ? 0) (((N41 ? 10) NIL) ((N42 ? 20) NIL))) ((N32 ? 0)
(((N43 ? 5) NIL))))) ((N22 ? 0) (((N33 ? 0) (((N44 ? -10) NIL))))))) ((N12 ? 0)
(((N23 ? 0) (((N34 ? 0) (((N45 ? 7) NIL) ((N46 ? 5) NIL))) ((N35 ? 0) (((N47 ? -20) NIL)))))
 ((N24 ? 0) (((N36 ? 0) (((N48 ? -7) NIL) ((N49 ? -5) NIL)))))))))

Horrible, isn't it?

And here's some code which doesn't work properly 😕 with some debugging print statements.

(defun dl-dfs (tree depth)
   "depth-limited depth-first traversal - create a list of nodes"
   (cond ((equal depth 0) nil)
             ((leafp tree)   (tree-node tree))
             (t       (format t "tree node = ~a" (tree-node tree)) (format t "~%")
                       (cons (tree-node tree)
                                  (dl-dfs-aux (tree-children tree) (1- depth))))
              ))

(defun dl-dfs-aux (tree-list depth)
   (if (null tree-list) nil
                 (cons (dl-dfs (car tree-list) depth)
                          (dl-dfs-aux (cdr tree-list) depth)
)))

---

(dl-dfs T0 5) produces:

tree node = (N0 ? 0)
tree node = (N11 ? 0)
tree node = (N21 ? 0)
tree node = (N31 ? 0)
tree node = (N32 ? 0)
tree node = (N22 ? 0)
tree node = (N33 ? 0)
tree node = (N12 ? 0)
tree node = (N23 ? 0)
tree node = (N34 ? 0)
tree node = (N35 ? 0)
tree node = (N24 ? 0)
tree node = (N36 ? 0)

((N0 ? 0) ((N11 ? 0) ((N21 ? 0) ((N31 ? 0) (N41 ? 10) (N42 ? 20))
((N32 ? 0) (N43 ? 5))) ((N22 ? 0) ((N33 ? 0) (N44 ? -10)))) ((N12 ? 0)
((N23 ? 0) ((N34 ? 0) (N45 ? 7) (N46 ? 5))
((N35 ? 0) (N47 ? -20))) ((N24 ? 0) ((N36 ? 0) (N48 ? -7) (N49 ? -5)))))

I wanted a straightforward list! *

---

* I'll spare you anything else - I have some new ideas for this afternoon. I already realised that the recursion is basically just reiterating the tree structure so the results either have to be flattened or I need to accumulate the nodes in a parameter as I go.

And I noticed the termination condition for depth = 0 is wrong, thank you.

---

Update: Wednesday 11th January 2017 - Here's the right answer.

(defun dl-dfs (tree depth)
  "Depth-limited depth-first traversal - create list of nodes"
  (flatten (dl-dfs1 tree depth)))

(defun dl-dfs1 (tree depth)
  "creates a list of nodes, but nested as a tree - needs to be flattened"
  (cond  ((or (leafp tree) (equal depth 0))  (tree-node tree))
         (t            (cons (tree-node tree)
                             (dl-dfs-aux (tree-children tree) (1- depth))))
         ))

(defun dl-dfs-aux (tree-list depth)
  (if (null tree-list) nil
    (cons (dl-dfs1 (car tree-list) depth)
          (dl-dfs-aux (cdr tree-list) depth)
          )))

(defun flatten (l)
  (cond ((null l)  nil)
        ((atom l) (list l))
        ((not (embedded-listp l)) (list l))
         (t       (append (flatten (car l)) (flatten (cdr l)) ) )
  ))

(defun embedded-listp (l)
   "If the list l contains a list, this returns a true, ie non-nil value"
   (member t (mapcar #'listp l)))

with result as follows: (dl-dfs T0 4)

((N0 ? 0) (N11 ? 0) (N21 ? 0) (N31 ? 0) (N41 ? 10) (N42 ? 20)
 (N32 ? 0) (N43 ? 5) (N22 ? 0) (N33 ? 0) (N44 ? -10) (N12 ? 0)
 (N23 ? 0) (N34 ? 0) (N45 ? 7) (N46 ? 5) (N35 ? 0) (N47 ? -20)
 (N24 ? 0) (N36 ? 0) (N48 ? -7) (N49 ? -5))

You won't care about this at all, but I'm quietly pleased. 😌